A few papyri transcribed

My work on Sappho has brought me in contact with a number of papyri, and I have found a number of images. This post is meant to contain all the transcriptions of said images which I did, so as to avoid overly long critical notes in some posts. We start by listing all the papyri that will be transcribed here, then for each papyrus we have the image (unless I am not allowed to publish it), and the transcription.

1. PSI XV 1470 r, consistent single fragment; contains commentary to and quote of the end of the Ode to Anactoria;
2. P.Oxy. 1232, big single two-col fragment; contains part of Hector and Andromacha;
3. P.Oxy. 2076, big single fragment; contains beginnings of final lines of Hector and Andromacha;
4. P.Oxy. 2288, single tall and narrow fragment; line beginnings of the Hymn to Aphrodite;
5. Ostrakon Florentinum (aka PSI 1300), single ostrakon; Idyll with Aphrodite;
6. PSI II 123, big single fragment; beginnings of 10 lines of Goddess Hera;
7. P.Oxy. 1231, fragment 1, plate II, column ii; more line beginnings for the previous poem; Grenfell-Hunt reading (image unavailable online);
8. P. Oxy. 2289 fr. 9; line middles for the previous papyrus;
9. Combining PSI 123, P.Oxy. 1231 fr. 1 plate II col. ii, and P.Oxy. 2289 fr. 9;
10. P.GC. inv. 105, fragment 2a, column ii, and fragment 2b, column i; line starts and line endings respectively for the previous poem (yet again) and of "second part" of "Dearest Offspring";
11. Adding this P.GC. to the combination above;
12. P.Oxy. 7, single big fragment; line endings for Cypris and Nereids;
13. P.Oxy. 2289, fr. 6; 2-3–letter portions of 3 lines from the above poem;
14. Cypris and Nereids without P.GC.;
15. P.GC. inv. 105, fr. 3 col. ii and fr. 5; line beginnings for the previous poem;
16. Adding P.GC. to Cypris and Nereids (which is now "divine Nereids");
17. P.GC. inv. 105, fr. 2a col. i; line endings for "Dearest Offspring" (aka Sappho 16);
18. The rest of P.GC. inv. 105, that is, fr. 1, fr. 3 col. i, fr. 4; bits of various fragments;
19. P. Oxy. 2289, frr. 4 and 10: more of the poem from P.GC. inv. 105 fr. 1 above, and a bonus;
20. Combining the above Oxyrhynchus fragment with the matching GC fragment;
21. P.Oxy. 1231, fr. 1, part of col. i: the biggest chunk of the pre-P.GC. text of Sappho 16;
22. P.Oxy. 1231, fr. 36: part of the poem above;
23. P.Oxy. 2166(a), fr. 2, two scraps giving more to the poem above;
24. Combining the Oxyrhynchus sources for this poem;
25. Throwing P.GC. into the mix;
26. P.Oxy. 1231, fr. 15: the so-called "Gongyla poem"; the paragraphs also includes the transcription of two more fragments, after a vain attempt at settling the transcription differences between Lobel-Page and Grenfell-Hunt without the image which was uploaded later; click here to get straight to the three transcriptions;
27. P.Oxy. 1231, fr. 56: "Let's sing for the couple!";
28. P.Oxy. 2166(a), fr. 6a: more to the above;
29. Combining sources for "Let's sing for the couple!";
30. P.Oxy. 1231 frr. 50-54 and P.Oxy. 2166(a) fr. 5 (two scraps) joined: a poem on nostalgia for youth;
31. Combining sources for the above poem;
32. P.Oxy. 1231 fr. 10: a poem on old age;
33. P.Oxy. 1231 fr. 13 + P.Oxy. 2166(a) fr. 7a: (potentially) another poem on old age;
34. Combination of the above;
35. P.Oxy. 1231 fr. 14: ode to a beautiful woman;;
36. P.Sapph. Obbink: a new poem, and a revolution to a P.Oxy. 1231 poem; apparently two big joined fragments, usually considered as a single fragment;
37. P.Oxy. 2289, fr. 5: traces of the first stanza of the first poem from above;
38. First poem from P.Sapph. Obbink: combined text;
39. P.Oxy. 1231, fr. 16: more to the second poem from above;
40. Second poem from P.Sapph. Obbink: combined text;
41. P.Oxy. 1787, frr. 1-2&2a: line ends for (how many?) poems;
42. P.Köln XI invv. 21351 and 21376: more to the above, and a doubt on the number of poems;
43. Combining P.Oxy. and P.Köln;
44. P.Berol. 9722, fol. 5 (LP p. 15): "O Atthis";
45. P.Berol. 9722, fol. 2 (LP p. 12): "I want to have died";
46. P.Berol. 9722, fol. 4 (LP p. 14): The other Gongyla poem;
47. P.Haun. 301: first part of the Kleis poem;
48. P. Mediol. ed. Vogliano: second part of the above;
49. P.Berol. 5006 recto: parts of a poem addressed to Charaxus;
50. P.Oxy. 424: more of that poem;
51. Combining the above;
52. P.Berol. 5006 verso: line endings of an otherwise non-extant poem;
53. P.Berol. 9722 fol. 3 (LP p. 14): Edmonds 84 vs. Lobel-Page 93 vs. my own reading;
54. P.Oxy. 1787 fr. 45: a book 4 colophon?;
55. P.Vindob. 29777a, two twoside fragments;
56. P.Oxy. 1787 frr. 35+40+XXI add. p. 136, three one-side fragments;
57. P.Oxy. 2293 frr. 1(a)+1(b): three (!) fragments which yield LP 90(1);
58. P.Oxy. 1231 fr. 19 + P.Oxy. 2166(a) fr. 4b: two fragments, a handful of letters;
59. P.Oxy. 1787 frr. 7(a) + 7(b) + 19 + LSM IV 9(a) + 9(b) + 9(c1) + 9(c2) + P.Oxy. XXI add. p. 135: a multitude of fragments with way too few numbers and decidedly arguable joints;
60. P.Berol. 9722, fol. 1 (LP p. 12): another crazy-faint parchment with wildly varying readings;
61. P. Köln inv. 0008r, big single fragment, not sure if Sappho but looks familiar;
62. P.GC. inv. 105 fr. 4 and P.Sapph. Obbink: to join or not to join?.

P.S.I. XV 1470: end of ode to Anactoria

I only consider the recto here, whereas the verso is not of interest (I don't even know if an image of it is found online). Here is the recto:


And here is my transcription, with completion in the middle and translation on the right.

Μ€ΟΒΟΝΒ̣ΟϹΟΙΛ[ ωΤωΝΚΑΙΟΤΡΟ[ ΤΟΥϹωΜ  ̣ΤΟϹΚ  ̣[ ΚΑΙΜ€ΤẠΤΑΥΤΑΤ[ ΦΗϹΙṆ ΧΛωΡΟΤ[ ]  ̣ΙΑϹ€ΜΜΙΤ€θ̣[ * ]ΛΙΓωΔ€ΠΙΔ€[ ]  ̣Ν̣ΟΜ̣€ΜΑΥ[ **

ἐ]μὲ ὁ βόνβ̣ος, ὁ ἴλ[ιγγος] ὤτων καὶ ὁ τρό[μος] τοῦ σώματος κα[τέχει] καὶ μετὰ̣ ταῦτα τ[άδε] φησίν̣· χλωροτ[έρα δὲ π]οίας ἔμμι, τεθ̣[νάκην δ' ὀ]λίγω δ' ἐπιδε[ύης φα]ίν̣ομ̣' ἐμαύ[τᾳ]

me the confusion, the sp[inning] of the ears and the tre[mor] of the body k[eeps] and after these things t[hese others] she says: green[er than gr]ass I am, and to not have d[ied] b[y l]ittle [I se]em to mys[elf]

Notes:
* The theta is uncertain because it doesn't close and so it looks like another epsilon.
** The nu is half cut off, and the mu has the left stroke vanished (faded, probably).

P.Oxy. 1232: first source for Hector and Andromacha

P.Oxy. 1232 is a group of three papyrus fragments. One of them is not related to the poem in the title, so I will not copy its transcription here. P.Oxy. X actually counts this one as joined with my fr. 2, though the image I have doesn't have that, and this says "the column breaks off", so I have no reason to believe these are actually a single 3-column fragment. Fragment 2 (coll ii-iii of fr. 1 in P.Oxy. X) is a very holey fragment of papyrus, of which I give an image from Wikimedia Commons below:

Disregarding (in part) what Grenfell and Hunt say about this, let me transcribe coll. ii-iii of P.Oxy. 1232 fr. 1, or what I see on the image.

Column 1:

ΚΥΠΡΟ ̣[      22-24 letters    ]ΑΙ   ̣ ΚΑΡΥΞΗ̣Λ̣θ̣[ . ]θ[    12 letters    ]€̣Λ̣[. . . .]θ̣€ΙϹ /ΙΔΑΟϹΤΑΔ€ΚΑ[. .]Ạ[. . .] . ΙϹ̣Τ̣ΑΧΥ̣Ϲ̣ΑΓΓ̣€̣ΛΟϹ      ΑΝ̣ω̣ ΤĀ́ϹΤÁΛΛΑϹΑϹΙΑϹ[. .]Δ̣€Ϲ̤ΑΝΚΛ€ΟϹ̣ΑΦθΙΤΟΝ· €ΚΤωΡΚΑΙϹΥΝ̣€ΤΑỊ . [ . ] ỊẠ . Ο . . €ΛΙΚωΠΙΔ̣Α̣· θΗΒΑϹ€ΞΙ€ΑΡΑϹΠΛΑΚΙΑϹΤΑ . [ .̄ .̈ ]ΝΑω ΑΒΡΑΝΑΝΔΡΟΜΑΧΑΝ€ΝΙΝ̣ΑΥϹΙΝ̣€ΠΑΛ̣Μ̣ΥΡ̣ΟΝ ΠΟΝΤΟ̣Ν̣·ΠΟΛΛΑΔ̣[. . .]Γ̣ΜΑΤΑΧΡȲ́ϹΙΑΚÄ́ΜΜΑΤΑ ΠΟΡΦΥΡ[ .̄ . . ]ΛΑΤΑÝΤ[. .]N̤ΑΠΟ̣Ι̤ Ḳ . ΛΑθΥΡΜΑΤΑ· ΑΡΓΥ . [. .]ΑΝ[. . .]θ̣[. . . . .] . [. . .]K̤ΑΛ€ΦĂΙϹ· ωϹ€ΙΠ̣ΟΤΡΑΛ€ωϹΔΑΝỌ . ΟΥϹ€̣Π̤ΑṬ[. .]ΦΙΛΟϹ· ΦΑΜΑΔΗ̤Λ̤θ€ΚΑΤΑΠΤ̣Ο̣ΛΙΝ€Υ . ỴΧ[. . .]ΝΦΙΛΟΙϹ· Ȧ . ΤΙΚΙΛΙΑΔΑΙϹ̣ΑΤ . Ν . . [ . ]ΥΠ€Υ̣ΤΡΟΧΟΙϹ̣ ΑΓ̣[ . ]ΝΑΙΜΙΟΝΟΙϹ·̣ . . [ . ]Β̤ΑΙΝ̣€Δ .́ ΠΑΙϹΟΧΛΟϹ Γ̣ . ΝΑΙ̤ ḲωΝTȦΜΑΠ[. .]θ̤€ΝỊ́ḲẠ . Ṭ . [. . .] . Ϲ̣Φ̣ΥΡω·̄ ΧῶΡΙϹ . ΑΥΠ€ΡΑΜΟΙΟθ . Γ̣[ . ] . P̤€Ϲ[ ΙΠΠ̣[. . .]Δ̣ . ΝΔΡ€ϹÝΠĀΓΟṆỴΠΑ̣Ρ̣[ Π[. . .] . ϹḤỊ[ . ]€ΟΙ·Μ€ΓÁΛω[ . ]ṬΙΔ[ Δ[. . . .] . ΑṆΙΟΧΟΙΦ̣[ Π̣[. . . . . .]ΞΑΓΟ[

Κυπρο ̣[      22-24 letters    ]ΑΙ   ̣ Κᾶρυξ ἦ̣λ̣θ̣[ε] θ[    12 letters    ]ε̣λ̣[. . . ἔ]θ̣εις /Ἴδαος τάδε κα[ῖν]α [φόρ]εις̣ τ̣άχυ̣ς̣ ἄγγ̣ε̣λος      ἄνω̣ Τᾶς τ' ἄλλας Ἀσίας [τά]δ̣' ἔσ̤αν κλέος̣ ἄφθιτον· Ἔκτωρ καὶ συνέται̣ρ[ο]ι̣ ἄ̣γοισ' ἐλικώπιδ̣α̣ Θήβας ἐξ ἰέρας Πλακίας τ' ἀπ' [ἀϊ]‹ν›νάω Ἄβραν Ἀνδρομάχαν ἔνι ν̣αῦσιν̣ ἐπ' ἄλ̣μ̣υρον Πόντο̣ν̣· πόλλα δ̣' [ἐλί]γ̣ματα χρύσια κἄμματα Πορφύρ[α κά]λα τ' αὖ τ[ρό]ν̤α πο̣ί̤κ̣ιλ' ἀθύρματα· Ἀργύρ[α τ'] ἀν[. . .]θ̣[. . . . .] . [. . .] κ̤ἀλέφαις· Ὠς εἶπ̣'· ὀτραλέως δ' ἀνό̣ρουσε̣ π̤άτ̣[ηρ] φίλος, Φάμα δ' ἦ̤λ̤θε κατὰ πτ̣ό̣λιν εὐρύ̣χ[ορο]ν φίλοις, Αὔτικ' Ἰλίαδαι σ̣ατίναι[ς] ὐπ' ἐϋ̣τρόχοις̣ Ἆγ̣[ο]ν αἰμιόνοις̣, ἐπ[έ]β̤αιν̣ε δὲ παῖς ὄχλος Γ̣υναί̣κ̣ων τ' ἄμα π[αρ]θ̤ενί̣κ̣α̣ν τ̣ . [. . .] . σ̣φ̣ύρω‹ν›· Χῶρις δ' αὖ Περαμοῖο θύγ̣[α]τ̣ρ̤ες [ Ἴππ̣[οις] δ̣' ἄνδρες ὔπαγον̣ ὐ̣π' ἄ̣ρ̣[ματα Π[. . .] . ς ἠ̣ΐ̣[θ]εοι, μεγάλω[σ]τ̣ι δ̣' [ Δ[. . . .] . ἀν̣ίοχοι φ[ Π̣̣[. . . . . ἔ]ξαγο[ν

Column 2:

[. . . . . . .]Φ . . . [. .]Ο[. . .] . €̣Δ̣€̣[. .] . [ . ] . [ [. . . . . . . .]ΙΚΑ̣Ϲ . . ΛΙΒΑΝΟ[ . ]ΤΟΝ€̣Δ̣€ΙΧΝΥ̣ΤΟ . Υ̣[ . ]ΑΙΚ̣€̣ . . [ . ]Λ . Λ̣ΥϹΔ[ . ]Ṇ Ξ [ . ] ΝỌϹΑΙΠΡΟΓ€Ν€́ϹΤ€ΡΑ[ ΠΑΝΤ€ϹΔ . ΝΔΡ[ . ]Ϲ[ .́ ]ΠΗΡΑΤΟΝĪΑΧΟΝΟΡθΙΟΝ ΠÁΟΝ’ΟΝΚΑΛ€Ο[ . ]Τ€Ϲ€ΚΑΒΟΛΟΝ€ΥΛÝΡĀΝ ΥΜ̣ΝΗΝΔ€ΚΤΟΡĀΚΑΝΔΡΟΜΑΧΑΝθ€ΟΙΚ€ . . [

[. . . . . . .]φ . . . [. .]ο[. . .] . ε̣δ̣ε̣[. .] . [ . ] . [ [. . . . . . . .]ικα̣σ . . λίβανό[ς] τ' ὀνε̣δ̣έχνυ̣το . υ̣[ . ]αικ̣ε . . [ὀ]λόλ̣υσδ[ο]ν̣ ὄ̣σαι προγενέστερα[ι Πάντες δ' ἄνδρ[ε]ς [ἐ]ήρατον ἴαχον ὄρθιον# Πάον' ὀνκαλέο[ν]τες ἐκάβολον εὐλύραν ῎Υμ̣νην δ' Ἔκτορα κἈνδρομάχαν θεοίκελο[ις

Some reading notes:

• That weird symbol on the left of the last two lines likely indicated the end of a poem and/or of a book of poems, probably the latter, seen that a «ϹΑΦ̣[  ]  Ϲ / ΜÉΛΗ̣[» is visible below column two’s end
• The doubtful C of ϹΑΤ . Ν . . [ . ] is a mere high stroke which could well be the top of a tau;
• The oblique dash and the ἄνω indicate an omission after l. 3;
• The superscript o in the raw transcription seems like an insertion: the writer of the papyrus had probably forgotten it, and a second writer added it between lines;

At first reading I thought ὀτραλέως had a chi instead of the lambda; then I zoomed in and saw that the ink only goes up left, not up right; if it were a chi, it should have both “legs” extending above the crossing point; since only one does, it must be a lambda, as Grenfell and Hunt stated; the image of this word is on right;

I know everyone reads a certain ΥΠĀΓΟΝ, but the half-letter that is visible is hardly a certain nu, with that incredibly horizontal stroke at the top, which really seems too horizontal to be a nu and not a gamma; I leave an uncertain nu partly to have a reading that makes sense, and partly because it is not impossible that this nu was weirdly shaped, e.g. with the slanted stroke being like an angle, that is bending sharply to reach the bottom in a short space with a steeper second part;

Fragment 3 is a smaller scrap which is only joined to fragment 2 because P.Oxy. 2076 «provides convincing supplements to four consecutive lines» on fragment 3, as this says, and also undoubtedly has the beginnings of the last lines of fragment 2. I cannot get ahold of an image to transcribe, which is why I'm just reporting Grenfell and Hunt's transcription from P.Oxy. X:

]Κ€ΛΟΙΘ€ΟΙ[ ]ΑΓΝΟΝΑΟΛ̣[ ]ΝΟΝ€ϹΙΛΙΟ[ ]ΤΟΝ€ΜΙΓΝΥ[ ]ωϹΔΑΡΑΠΑΡ̣[ ]Ν€Δ€Ϲ̣ . . [

ἴ]κελοι θέοι[ς ἄ]γνον ἀόλ̣[λεες ]νον ἐς Ἴλιο[ν ]τον ἐμίγνυ[σαν ]ως δ' ἄρα πάρ̣[θενοι ]νε δ' ἐς̣ . . [

P.Oxy. 2076: more of Hector and Andromacha

Again I have an image of this, and this time there is no Grenfell-Hunt transcription for comparison, because this is in volume XVII which is not on archive.org (or wasn't the last time I checked, when I downloaded XV as the last available one), so I transcribed it blind. Here's the image:

Which I transcribed as:

. Ρ̣ΜΑΤA̤ . [ ΑΥ̣ΛΟϹΔΑ̣Δ̣Y̤[ . ]€̣ΛΗ[ ΚΑΙΨ[ . ]ΦΟ[. .]ΡΟΤΑΛ[ Α€ΙΔΟΝΜ€ΛΟϹΑΓ . [ ᾹΧωθ€ϹΠ€̣ϹỊ[ . ]Γ€Λ̤[ ΠΑΝΤΑ͞ΙΔΗϹΚΑΤÓΔΟ[ ΚΡΑΤΗΡ€ϹΦΙΑΛΑ̣ỊP̤Ο̣[ ΜΥΡΡΑΚΑΙΚΑϹΙΑΛΙΒ[ ΓΥΝΑΙΚ€ϹΔΟΛO̤ΛΥΖΟ̣[ ΠΑΝT̤€ϹΔΑΝΔΡ€Ϲ€̣[ ΠÁΟN̤ΟΝΚΑΛ€ΟΝΤ€Ϲ[ ΥΜΝΗΝΔ€ΚΤΟΡΑΚΑ[                                    ϹΑΠΦΟ[                                                 Β                                    ΜΙΤ̣I̤ΝΥΔΥ[

Ὄρ̣ματα̤ι [ Αὖ̣λος δ' ἀ̣δ̣υ̤[μ]έ̣λη[ς Καὶ ψ[ό]φο[ς κ]ροτάλ[ων Ἄειδον μέλος ἄγν[ον Ἄχω θεσπε̣σί̣[α] γέλ̤[ος Πάντᾳ δ' ἦς κὰτ ὄδο[ις Κράτηρες φίαλα̣ι̣ ρ̤ο̣[ Μύρρα καὶ κασία λιβ[ Γύναικες δ' ὀλό̤λυζο̣[ν Πάντ̤̣ες δ' ἄνδρες ἐ[ Πάον̤ ὀνκαλέοντες [ ῎Υμνην δ' Ἔκτορα κα[                                    Σαπφο[ῦς                                                 β                                    μιτ̣ι̤νυδυ[

Actually, after the colophon, we have an obscure fragment which reads ] . ΑΡ€ΦΙΛ€ΙΔΥ[ or ] γὰρ ἐφίλει δυ[. In my defense, the alpha and the rho are linked and look like a mu, the epsilon is two strokes, one almost vertical (my iota), and one horizontal (the top of my tau) which crosses the circle of the phi out, the lambda is one slanted stroke plus something joining into the bottom of the epsilon (hence my nu), the rest of the epsilon is attached to the iota and makes it look like an upsilon. To make all that more clear, below I show you how this was supposed to be read (left) and how I read it (right). The dotted orange line is a sign that's basically gone.

P.Oxy. 2288: line beginnings of the Hymn to Aphrodite

This is a very narrow fragment with the beginnings of most lines of the Hymn to Aphrodite, with a few letters taken away from the very beginnings. Here is the image.

And here is the transcription:

]ΙΚΙΛ  ̣  ̣  ̣ ̤  ̣Λ̤  ̣́  ̣ΛΟ̤  ̣ {...}ẠṂṆẠ {....}Αθ̣Υ{.} ẠΤÝΙΔ  ̣ Ϲ€̄ṂΑϹΑÝ Ỵ€ϹΠẠΤΡ  ̣  ̣ΥϹΙΟΝΗ̤Λ̤  ̣ ΜΥΠ̣[.]ϹΔ€ Κ̣€̣€Ϲ̤ϹΤΡΟ᷍Υ ḲΝΑΔĬΝ̤ ỌϹΔΙΑΜ€̣Ϲ̣  ̣  ̣   ̣ΑΔ€ξÏḲΟ   ̣ΗΔΙΑϹẠ[.. Ρ̤€ΟΤΤΙ  ̣[.]  ̣ ΥΤ€́{.}A̤{.}[.]M̤ ΟΤΤÍ{.}[.]{.} A̤I̤N̤  ̣ΛΑI̤ {.}C̣ĂΓΗṆ ]ΠΦ[

Πο]ίκιλ . . . [ ] . λ̤ . . λο̤ . [ ]{.. δ}ά̣μ̣ν̣α̣ [ ]{....}αθ̣υ{.} [ ]α̣ τυίδε [ Τᾶ]ς ἔμ̣ας αὔ[δως Ἔκλ]υ̣ες πά̣τρο[ς Χ]ρσιον ἦ̤λ̤[θες Ἄρ]μ' ὐπ̣[α]σδε[ Ὤ]κ̣ε̣ες̤ στροῦ[θοι Πύ]κ̣να δῐν̤[ ]ο̣ς διὰ μέ̣σ̣δω ] . α δ' ἐξίκ̣ο[ντο Μηδιασα̣[ισ' Ἤ]ρ̤ε' ὄττι . [ . ] . [ Δη]ὖτε {.}.α̤{.}[.]μ̤[ ] ὄττι {.}[.]{.}[ Μ]α̤ι̤ν̤όλαι̤ [ ]{.}ς̣ ἄγην̣ [ Ψά]πφ[

What does this give us? Here is my take:

• First of all, we address l. 19. My transcription above is assuming we have a compressed sigma (hence doubtful) and a half-nu (hence doubtful), and "{.}" denotes a blank space which could fit a letter. The "vulgate" reads a psi (certain according to Campbell) in there. But where? The best I can do is the following, where the left side is just the line, and the right side has the near-invisible psi redrawn:
  
  
  Now, that's hardly certain for me. As I said on Latin SE, «I for one believe this is far too close to the sigma to actually be what was written there. There is something like a mini-psi far down in a sort of subscript placement, but that can hardly (IMHO) be considered an actual letter. So I'm very prone to discarding this psi idea. For other reasons, M.L. West, a very famous linguist I only found out about when I found that link, discards it. This reading, however, is accepted by many (Greek Wikisource, BA, Campbell, these guys).». More precisely, this is how M.L. West discards the psi:
  
  
• In l. 9, the zeta of the codices is unambiguously dispelled by the clear σδ cluster.
• Edmonds' idea of a dual at ll. 9-10 is dispelled by the two sigmas in l. 10, which say plural, though one is near-invisible, but there is a trace.
• The possibility of μειδιάσαισ' at l. 14 is dispelled by the certain eta.

Ostr. Flor., aka P.S.I. XIII 1300: the Idyll with Aphrodite

This is an ostrakon, that is, a piece of terracotta, where a scribe with an IMHO appallingly awful handwriting has written down the "full" ode, of which only pieces were previously known by indirect quotation. Here is the image:



And here comes my transcription, with a few (more like a lot of) comments, and tweaked perhaps after comparison with what these guys have:

. . P̣ẠΝΟθ€ΝΚΑΙΤΙΟΥ Δ̤ €̣ΥΡΥΜΜ€ΚΡḤṬẠC̤€̤Π ΝΑΥΓΟΝΑΓΝΟΝΟΠ ΑΡΙ€ΝΜ€ΝΑΛϹΟϹΜΑΛỊ ωΜΟΙΔ€ΜΙθΥΜΙΑΜΚ̣Ν̣ΟỊ ΑΝωΤω·€ΝΔΥΔΩΡΨΥΧΡOṆ ̣ ̣ Ν̤ΜΟΙΔΥϹΔωΝΜΑΛΙΑΝ Β̤ΡΟΔΟIϹΔ€ΠΑ̤ΙϹΟΧωΡΟϹΙΑϹΤ̤Α O̤ϹΤ̤Α̤ΙθΥϹϹΟΜ€ΝωΙΔ€ΦΥΛΛω ω . ΑΚΑΝ̣Α̤ΥΡΡ̣ΟΝ€ΝΔ€Λ€ΙΜω ΙΠ̣Π̣Ο̣Β̣ΟΤΟϹΤ€θΑΑΛ€ΗΝΟỊϹ̣̣ṬΝ̤Α̣IΝ̣ ΝθΙϹΑΝθ€ϹΙΝΑΙΔΑΙΗΤΑΙΜ€ΛΛΙ ΧΑΠΝ€̤Ο̤ΙϹΙΝ̣      €ΝθΑΔΗϹΥϹΤ€Μ . Λ̣Ο̣ΙϹ̣Α̣ΚΥΠΡΙΧΡΥϹΙΑΙϹΙΝ€Ν Κ̣ΥΛ . Ḳ . ϹϹΙΝΑΒ̣ΡωϹΟΜΜ̣€Ι Χ̤Μ̤€̤Ν̣Ο̤ΝθΑΛ̣ỊΑΙϹ̣ΙΝ̣Ν€ . . . Α̣Ρ̤ . Α̤Ρ̣ωΝΟΧ̤ΟΑΙϹỌ .

. . ρ̣ά̣νοθεν κα{ι}τίου δ̤ε̣ῦρύ {μ}μ' ἐκ Κρή̣τα̣̣ς̤ ἐ̤π[ὶ ναῦ{γ}ον ἄγνον ὄπ[πᾳ χ]άριεν μὲν ἄλσος μαλί[νων β]ῶμοι δ' ἔ‹ν›ι θυμιάμ‹ε›ν̣οι̣ λιβ]ανώτῳ· ἐν δ' ὔδωρ ψῦχρον̣ [ ̣ ̣ ]ν̤ μοι δ' ὔσδων μαλίαν β̤ρόδοις δὲ πα̤ῖς ὀ χῶρος ἴαστ̤α[ι ]ο̤στ̤' α̤ἰθυσσομένω‹ν› δὲ φύλλω[ν κ]ῶμα κάν̣α̤υρρ̣ον· ἐν δὲ λείμω[ν ἰπ̣π̣ό̣β̣οτος τέθαλε ἠ‹ρί›νοι̣σ‹ι›ν̤ ἄ{ι}ν[ νθις ἄνθεσιν αἰ δ' ἄ{ι}ηται μέλλι- -χα πνέ̤ο̤ισιν      ἔνθα δὴ σὺ στέμ[- ἔλ̣ο̣ισ̣α̣ Κύπρι χρυσίαισιν ἐν κ̣υλίκ̣εσσιν ἄβ̣ρως ὀμμ̣εί[ -χ̤μ̤ε̤ν̣ο̤ν θαλ̣ί̣αισ̣ιν̣ νέ[ . . . α̣ρ̤ τα̤ρ̣ ὠνοχ̤όαισο̤ν

This mess, of course, calls for a frickton of comments.

• First and foremost, the meter isn't respected at all, either because the ostrakon had the shape it had and didn't allow linebreaking consistent with the meter, or because, as SaffoComm1 suggests, this was a dictation exercise and the scribe didn't know this was a poem, or both; whatever the case, this also entail words broken into two, as happens with στέμ[ματ', ὀμμείχμενον, νέκταρ.
• Next, the terrible beginning. As you will see from images below, "ΑΝΟΘ€Ν" is clearly visible. Before that, I see potentially three letter vestiges. One way I could retrace those letters is:
  
  
  Those who read it like this will have a very holey stanza at the beginning. I fell into this "trap" long ago, and partially completed that stanza. However, I am now persuaded that the actual text was:
  
  
  That is because I am persuaded with the hypothesis that the scribe, under dictation, forgot that part in stanza 3, then noticed the forgetfulness, wrote the missing bit up top, and added "ἄνω" ("above") or similar somewhere near the missing bit's place, and that word was lost due to ink fading - look at how faint some letters are! Some experts explain this by saying the scribe was writing from memory, and made a memory error, or had a hole in his memory which is the hole in stanza 3, completed up top, but I prefer the dictation idea because all the repetitions, which will be discussed below, can be explained away as corrections by the teacher or by the scribe themself. Whatever the case, this reading is also favored by the fact that, while retracing attempt 2, I noticed the vertical line of the kappa seems barely visible, whereas I previously had not seen any trace of it.
• The KAITIOY is pretty undoubtable. However, the only possible interpretation I can give it is καὶ τί οὐ (or καί τι οὐ), "and what not?" (or "and something not"), which doesn't really fit the third stanza, and besides, with ΑΝΟΘ€Ν so clearly giving us a "from" complement just before, we'd expect a verb of motion, like κατίου, "come down". That is the reason for my emendation.
• I know people read ναῦον with no emendations, but you can't really mark a dropped letter in a critical edition, and that gamma is pretty evident:
  
  
  
  Unfortunately, ναῦγον yields nothing from my references, which is why I amended it to the usual ναῦον.
• Now for the θΥΜΙΑΜΚ̣Ν̣Ο part at the end of l. 5. Here's the image of ΑΜΚ̣Ν̣Ο:
  
  
  
  Oh my gosh. Now I see the epsilon. I previously read that as IC, then today I looked at it and saw a near-vertical stroke and a curved stroke which could add up to a K or IC, and now, looking at the preview of this post, I see how they could, in fact, add up to the weirdest-looking epsilon ever. This handwriting is just… Man, if you're writing a lunate €, you should do a C stroke and add a dash in the middle, not pretend like you're doing a reversed 3 and do two C strokes connected to each other! So we finally get that epsilon we wanted to have the text usually printed.
• Now for the beginning of l. 8:
  
  
  
  From MOI on, things are pretty clear. What happens before is a faint-ass mess. I believe the probable reading is ΔΑΙ, so maybe the MOI is an error and the mess is a correction by the teacher? Of course, the teacher seems to have forgotten to correct ΔΥϹΔωΝ to ΔΙΥϹΔωΝ, but that is a small mistake, so maybe the teacher missed it?
• The end of line 9 is another problem. Image:
  
  
  Now, the χωροσιασ part is pretty clear, the tau is just a spot on top of the line, probably falling down to touch the bottom of the uncertain alpha, and the iota is clear enough. So this gives ἴασται, if this is one word, and who knows what otherwise. However, that word can only be taken as a psilosis from ἴασθαι, probable Aeolic form of ἰῆσθαι, meaning "is cured", but really, the whole place is cured by roses? What? Let us also look at the beginning of the following line:
  
  
  
  The more immediate reading of this would be OCINI. However, the first iota is connected to the nu, so one could take that iota as a tau, the connection being the top, and then the nu could be broken into an alpha and a connection of that alpha to the following iota. Another way of obtaining this reading is to observe that this scribe has his own stupid way of writing taus, which I will metion in the next sentence, and lets us take the iota, the connection, and the first half-nu as the tau, and the rest of the nu as the alpha. The way I mentioned is to trace half the horizontal stroke, then go down with the vertical one, then back up to do the rest of the horizontal one, whereas the intelligent (and non-confusing for readers) way to trace it is –, then | (or viceversa | then –). However you choose to obtain it, this is what I chose for the above, because AI gives me a complete αἰθυσσομένων. The rest is OCT', and could be taken as a correction by the teacher indicating the verb is supposed to be elided. The omicron could be taken as an alpha, since it is so faint and almost just a blot. Recognizing a quotation by Athenaeus, we want φύλλων to end a Sapphic endecasyllabic line, so ]ίαστ' would need an extra long syllable before it, e.g. the typically printed ἐσκίαστ', "is shadowed".
• The next problem is in l. 10. Image:
  
  
  
  The beginning KA and the ending ON are undoubted. The middle is a million of possibilities. The first reading I gave was ΚΑΝΥΡΡΟΝ, and this is the most immediate: see the nu, the upsilon (with a faint left slanted stroke), the two rhos. We have to take the matter of taus (cfr. previous note) into account, making the nu a possible tau. Attached to this nu/tau is a blot which could be an alpha or an omicron. The second rho has such a tight top that it could just be a blot on top of an iota. So how many readings can we get? ΚΑΝΥΡΡΟΝ, ΚΑΤΥΡΡΟΝ, ΚΑΝΥΡΙΟΝ, ΚΑΤΥΡΙΟΝ, ΚΑΝΑΥΡΡΟΝ, ΚΑΤΑΥΡΡΟΝ, ΚΑΝΑΥΡΙΟΝ, ΚΑΤΑΥΡΙΟΝ. Well, that's not all: the upsilon could be a gamma, with that faint stroke being a connection. So we have a grand total of 16 possible readings. Does any of these yield something meaningful? Not according to perseus. Great! Locus desperatus, I guess. Most likely reading? Well, the nu has a strange shape for a nu, so I'll read a tau there. The upsilon still looks more like an upsilon than a gamma. As for the alpha… bracket it. So ΚΑΤ(Α)ΥΡΙΟΝ, because indeed the second rho is more an iota than a rho. Actually, if we read κατ' αὔριον, we could make it mean something. So I'll go with that. Let's just see what Bonnaria thinks. Κατάγρει? «Take the sleepiness of the shaking leaves»? After all those present tenses describing the place, with more presents in the following stanza, you give me an imperative? SMH. No way. I trust Athenaeus. Or at most I imply ἐστι and read this as κάτερρον. [Checkup-time me cannot figure out what that word was supposed to mean.]
• Now the ΗΝΟỊϹ̣̣ṬΝ̤Α̣IΝ̣ / ΝθΙϹΑΝθ€ϹΙΝ around the boundary between ll. 11-12. Images (end of l. 11 on the left, beginning of l. 12 on the right):
  
  
  
  There is not much to be doubted about the reading of the right-hand image: at most we could read Ν€ΙϹ instead of ΝθΙϹ. As for the left-hand image, ΑΛ€ΗΝ is clear, then we have a scribble, then a pretty clear AIN, with at most a doubtful alpha. What IS that scribble though? Let's zoom in:
  
  
  
  It looks like a little wave going up, down, up, down, up, not so far down, up, down, up, down, and then back up with an unusual angle at the bottom. Oh I included the HN in the above image. Oops :). The first two up-down's and the following up could be two taus, then the down-up could be an upsilon, and the rest could be a nu. So HNTTYN. Unreadable. Well, the first up is a bit blotty, maybe an omicron connecting to the tau. So HNOTTYN. No meaning. Blotty up = O, down-up = upsilon, down-up lost a curving bottom for a C, down-up = upsilon again, then a nu for the rest. HNOYCYN. ἦν οὐ σύν? Was not with… flowers (ἄνθεσιν)? How does it link to the previous part of the poem, with that undoubtful τέθαλε? That second upsilon, and the first one too, are a bit too curvy though, compared to others. Maybe the scribe decided to link those two iotas in such a confusing way to confuse the ever-living frick out of us? Probably. So this would be… ἤνοισιν. What? Hypothizing an Aeolic eta for ει, εἴνοις would be nonsensical. So we need to amend this. We expect an adjective, and the meter suggests an iambic meter, since it's the start of a line. Hence ἠ‹ρί›νοισιν presents itself. But we still have the αιν/νθισανθεσιν jumble. Well, ἄνθεσιν fits perfectly, and the rest seems spurious. Probably either a scribal error where the scribe wrote nonsense, then went WTF and wrote the correct version, or the teacher getting stuck on reading and the scribe deciding to record the teacher's error.
• Let's see the beginning of l. 16:
  
  
  
  Pretty far-fetched, I know. That's why all the uncertainty.
• Let's finally look at l. 17:
  
  
  
  Let me retrace that for you, indicating areas of untraceable vestiges with rectangles:
  
  
  
  The end is fairly certain, modulo having a monster form (says Braun according to Bonnaria) and a bad misspelling. I'll take the Bonnaria choice form οἰνοχόεσσον as an amendment and close that case. The beginning is another case. Ν€ at the end of the previous line, together with ΑΡ retraced here, suggest νέκταρ, to which Athenaeus agrees. This is why I gave an underlined tau before the AP, and now I saw what could be a trace of the up-down manner of writing it that this scribe employed. What comes before (. . . AP) is hard to explain. The very faint traces suggest N/M, dot, I, then this AP. Maybe the scribe wrote Ν€ / ΤΑΡ and the teacher corrected it on a little space left before TAP by the scribe? I'll just read νέ[κ]ταρ there and call it a day.

Having said all that, let me reinstate proper line divisions, drop all spurious letters, and complete where possible:

δ̤ε̣ῦρύ μ' ἐκ Κρή̣τα̣̣ς̤ ἐ̤π[ὶ τόνδε] ναῦον
ἄγνον, ὄπ[πᾳ δὴ χ]άριεν μὲν ἄλσος
μαλί[νων, β]ῶμοι δ' ἔ‹ν›ι θυμιάμε̤-
ν̣οι̣ λιβ]ανώτῳ·

ἐν δ' ὔδωρ ψῦχρον̣ †[ ̣ ̣ ]ν̤ μοι† δ‹ι›' ὔσδων
μαλίαν, β̤ρόδοισ‹ι› δὲ πα̤ῖς ὀ χῶρος
‹ἐσκ›ίαστ̤', α̤ἰθυσσομένω‹ν› δὲ φύλλω[ν
κ]ῶμα κατ̣ά̤‹ρρει›·

ἐν δὲ λείμω[ν] ἰπ̣π̣ό̣β̣οτος τέθαλε
ἠ‹ρί›νοι̣σι̣ν̤ ἄνθεσιν, αἰ δ' ἄηται
μέλλιχα πνέ̤ο̤ισιν [
     ]

ἔνθα δὴ σὺ στέμ[ματ'] ἔλ̣ο̣ισ̣α̣, Κύπρι,
χρυσίαισιν ἐν κ̣υλίκ̣εσσιν ἄβ̣ρως
ὀμ‹με›μ̣είχ̤μ̤ε̤ν̣ο̤ν θαλ̣ί̣αισ̣ιν̣ νέ[κ]τα̤ρ̣
‹οἰ›νοχ̤ό‹εσ›σο̤ν.


Of course, stanza 3 should actually end:

μέλλιχα πνέ̤ο̤ισι{ν} [φύτων] καράνο-
θεν κατίο‹ι›[σαι.]

However, the above transcription must go into the post on the poem, where I have to discuss my way to the text I translated, which fell into the super-holey stanza trap, so I cannot fill the lacuna with the top in that post.

Just to give you an idea of just how horrible the handwriting is, I give you the transcription of this same ostrakon found in Lobel-Page, commenting that, while in some places I cannot agree with it, in a few I can see both my reading and theirs now that I've seen theirs.

. . P̣ẠΝΟθ€ΝΚΑΙΤΙΟΥ Δ̤ €̣ΥΡΥΜΜ€ΚΡḤṬẠC̤€̤Π ΝΑΥΓΟΝΑΓΝΟΝΟΠ ΑΡΙ€ΝΜ€ΝΑΛϹΟϹΜΑΛỊ ωΜΟΙΔ€ΜΙθΥΜΙΑΜΚ̣Ν̣ΟỊ ΑΝωΤω·€ΝΔΥΔΩΡΨΥΧΡOṆ ̣ ̣ Ν̤ΜΟΙΔΥϹΔωΝΜΑΛΙΑΝ Β̤ΡΟΔΟIϹΔ€ΠΑ̤ΙϹΟΧωΡΟϹΙΑϹΤ̤Α O̤ϹΤ̤Α̤ΙθΥϹϹΟΜ€ΝωΙΔ€ΦΥΛΛω ω . ΑΚΑΝ̣Α̤ΥΡΡ̣ΟΝ€ΝΔ€Λ€ΙΜω ΙΠ̣Π̣Ο̣Β̣ΟΤΟϹΤ€θΑΑΛ€ΗΝΟỊϹ̣̣ṬΝ̤Α̣IΝ̣ ΝθΙϹΑΝθ€ϹΙΝΑΙΔΑΙΗΤΑΙΜ€ΛΛΙ ΧΑΠΝ€̤Ο̤ΙϹΙΝ̣      €ΝθΑΔΗϹΥϹΤ€Μ . Λ̣Ο̣ΙϹ̣Α̣ΚΥΠΡΙΧΡΥϹΙΑΙϹΙΝ€Ν Κ̣ΥΛ . Ḳ . ϹϹΙΝΑΒ̣ΡωϹΟΜΜ̣€Ι Χ̤Μ̤€̤Ν̣Ο̤ΝθΑΛ̣ỊΑΙϹ̣ΙΝ̣Ν€ . . . Α̣Ρ̤ . Α̤Ρ̣ωΝΟΧ̤ΟΑΙϹỌ .

. Ρ̣Α̣ΝΟΘ€ΝΚΑΤΙΟΥ Δ€ΥΡΥΜΜ̣€ΚΚΡΗΤΑ̣Ϲ̣ . Π̣ . ΝΑΥΓΟΝΑΓΝΟΝΟΠΠ̣ ΧΑΡΙ€ΝΜ€ΝΑΛϹΟϹΜΑΛΙ . ωΜΟΙΔ€ΜΙΘΥΜΙΑΜ€ΝΟΙ . ΑΝωΤω€ΝΤΥΔωΡΨΥΧΡΟ . . Λ̣Α̣Τ̣ΙΔ̣ΙΔΥϹ̣Χ̣ωΝΜΑΛΙΑΝ Β̣ΡΟΤΟΙϹΟΤ€Π̣€̣ϹΟΧωΡΟϹΚΙϹΚΙ ΑϹΤΑ̣ΙΘΥϹϹΟΜ€ΝωΝΔ€ΦΥΛΛωΝ ΚωΜΑΚΑΤΑΓΡΙΟΝ€ΝΔ€Λ€ΙΜωΝ ΙΠΠΟΒΟΤΟϹΤ€ΘΑΛ€Τω̣Τ . . . Ι̣ΡΙΝ ΝΟΙϹΑΝΘ€ϹΙΝΑΙΑΑΙΗΤΑΙΜ€ΛΛΙ ΧΑΠΝ€̣Ο̣ΙϹΙΝ€ΝΘΑΔΗϹΥϹ . €Μ €ΛΟΙϹΑΚΥΠΡΙΧΡΥϹ€ΑΙϹ€ΝΚΥ ΛΙΚ€̣ϹϹΙΝΑΚΡωϹ . . Μ€ Χ̣Μ̣€̣ΝΟΝΘΑΛΙΑΙϹΟΝ . . Κ . . ΡωΝΟΧΟΑΙϹΟΝ

P.S.I. 123: beginnings of 10 lines of Goddess Hera

Here comes the image:



The first line is too mutilated to read anything sensible out of it, though there are traces of ink. Transcription of the rest:

Τ€ΞΑΔΟΚΗ ΠΛΑϹΙΟΝΔΗΜ ΠΟΤΝΙΗΡΑ·ϹΑΧ ΤΑΝΑ€ΡẮΤΑΝΑΤΡ ΤΟΙΒΑϹΙΛΗ€Ϲ €ΚΤ€Λ€ϹϹΑΝΤ€Ϲ̣ ΠΡωΤΑΜ€ΝΠ . ΤΥΪΔ'ΑΠΟΡΜᾹ́θ€̣ ΟΥΚ€ΔΥΝΑΝΤΟ ΠΡΙΝθϹ€ΚΑΙΔῘᾸΝT̤ ΚΑΙθΥώΝΑϹΙΜ .

τ' ἐξ ἀδοκή[τω Πλάσιον δή μ[ Πότνι' Ἦρα, σὰ χ[ Τὰν ἀράταν Ἀτρ[εΐδαι -τοι βασίληες Ἐκτελέσσαντες̣ Πρῶτα μὲν π . [ Τυίδ' ἀπορμάθε̣[ντες Οὐκ ἐδύναντο Πρὶν σὲ καὶ Δί' ἀντ̤[ Καὶ Θυώνας ἰμ . [

A couple notes:

• The first line has a few pseudo-doubts: the epsilon is faint and hard to see, the alpha is topless and the eta has the right vertical stroke cut away, but they leave no doubts that they are an epsilon, an alpha and an eta. Also, that line is from another poem, as the fancy symbol to the left of the letters denotes. The underlining of the tau is probably actually a part of the fancy symbol. To the left of said symbol, τῶμον seems to be read. Not sure what it was supposed to mean.
• The correction in l. 3 is believed to be misguided by experts, and indeed meter demands the PA syllable be long, whereas with the correction it is short. The corrector probably didn't acknowledge the meter and mistakenly assumed the word there was ἐράταν, which indeed has a short PA syllable, as the corrector probably also marked.
• The tau in l. 6 is bottomless, but the only other option is a gamma, which might be a thing if the scribe traced it in two strokes, but I believe gammas were usually traced in one bending stroke, which would rule out a gamma here since the top stroke is on both sides of the vertical one.
• The letters at the ends of ll. 6-7 are very mutilated. I decided to play it safe in l. 6, though a round stroke would suggest a very doubtful epsilon, whereas I pushed it a bit in l. 7 where you see a round stroke and a trace of a horizontal one, which would spell epsilon, hence the singly-underdotted epsilon.
• The tau at the end of l. 9 has only left a trace of its horizontal stroke, hence the double underdot.
• No idea what the X on the left of l. 10 means.

P.Oxy. 1231, fragment 1, plate II, column ii: more line beginnings

I have no idea what this looks like, since no image is online. For notes, I refer you to Grenfell-Hunt P.Oxy. X, where a couple comments are made on PSI too. Here is the transcription found in that book, dropping the t' ἐξ ἀδοκή[τω line from above (RHS column mine):

ΠΛΑ̣Ϲ̣[ ΠΟΤ[ . ]ỊΗ̣[ ΤΑΝΑΡΑΤ[ ΤΟΙΒΑϹΙΛ[ €ΚΤ€Λ€Ϲ[ ΠΡωΤΑΜ[ ΤỴ́[ . ]Δ̣ΑΠΟ[ ΟΥΚ€ΔΥΝ[ ΠΡΙΝϹ€̀[ ΚΑΙθΥωΝ[ ΝΥΝΔ€Κ ΚÀΤΤΟΠΑ[ ΑΓΝΑΚΑΙΚΑ̣[ [ . ]ΑΡΘ[ [ . ]Μ̣ΦΙϹ̣[ [] [] . . Α̣Ν̣ΙΛ[ €ΜΜ€Ν[ Ρ̣ΑΠΙ[

Πλά̣σ̣[ιον Πότ[ν]ι̣' Ἦ̣[ρα Τὰν ἀράτ[αν Τοι βασίλ[ηες Ἐκτελέσ[σαντες Πρῶτα μ[ὲν Τυ̣[ί]δ̣' ἀπο[ Οὐκ ἐδύν[αντο Πρὶν σὲ Καὶ Θυών[ας Νῦν δὲ κ[ Κὰτ τὸ πά[λαιον Ἄγνα καὶ κα̣[λ [Π]αρθ[εν [Ἀ]μ̣φὶ σ̣[ [] [] . . α̣ν̣ιλ[[ Ἔμμεν[ ρ̣απι[

The only comment I can make is that the two lost lines may have been either 2, 6, or 10, by meter, and assuming this poem had no more than 7 stanzas, which is the maximum attested for Sappho poems.

P.Oxy. 2289 fr. 9: middle bits of ll. 5-8 of the previous poem

This fragment undoubtedly contains portions of lines from the previous papyrus: the overlap is overwhelmingly big, this conclusion is just inevitable. It's part of a group of 36ish fragments called P.Oxy. 2289, put together for I don't know what reason. Here's the image:

This is my reading:

Η̤€̤[ . ] ΝΤ€ϹΜ̣ . ΝΠ€Ρ . ΜÁθ€Ν̣ Ν .

]η̤ε̤[ς] ]ντες μ̣[ μ]ὲν πὲρ [ ορ]μάθεν̣̣[τες ]ν .

A couple notes:

• The mu in l. 2 is cut by the margin, and it could be a nu.
• You can actually see a trace of an epsilon at the start of l. 3.
• The vestige at the end of l. 3 is mysterious: it looks like it has a vertical stroke, which suggests an iota, but then there is the horizontal one, which suggests an epsilon. If epsilons looked like capital Es, it would be one no doubt, but epsilons are lunate €, so that vertical stroke goes too high up for one.
• The mu at the start of l. 4 is half-cut but couldn't be anything else.
• The nu at the end could be a mu.

Combining PSI 123, P.Oxy. 1231 fr. 1 plate II col. ii, and P.Oxy. 2289 fr. 9

So let us combine these papyri for a single text, coloring P.Oxy. 1231 yellow and P.Oxy. 2289 blue, reading the weird vestige of 2289 as an iota:

Πλάσιον δή μ[
Πότνι' Ἦρα, σὰ χ[
Τὰν ἀράταν Ἀτρ[εΐδαι
-τοι βασίληες
Ἐκτελέσσαντες μ̣[
Πρῶτα μὲν πὲρ Ἴ̤[λιον
Τυίδ' ἀπορμάθεν̣̣[τες
Οὐκ ἐδύναντο
Πρὶν σὲ καὶ Δί' ἀντ̤[
Καὶ Θυώνας ἰμ . [
Νῦν δὲ κ[
Κὰτ τὸ πά[λαιον
Ἄγνα καὶ κα̣[λ
[Π]αρθ[εν
[Ἀ]μ̣φὶ σ̣[
[]
[]
. . α̣ν̣ιλ[[
Ἔμμεν[
ρ̣απι[

P.GC. inv. 105, fragment 2a, col. ii: more line beginnings; and fragment 2b, column i: line endings

I know these are usually numbered 2 together, but they are separated, how can I consider them a single fragment if there is literally nothing joining them? I mean, one could even be skeptical about whether or not they should be beginnings and ends of the same lines! I go with that theory since some supplements they provide are pretty convincing and we know 2a and 2b come from the same "box" (yes, thy were used as packaging material) so we need to place them in some order, but from this to a single fragment number seems an overly big step. Anyways, I can't put the image up because it's in a book and it's copyrighted, so I will only put up certain bits in the reading notes. The first 7 lines are from the "second part" of Dearest Offspring, while the rest is from Goddess Hera above. That said, here comes the transcription, organized as two pairs of columns: left pair is the raw transcription, right pair is the interpretation, both with 2a col. ii on the left and 2b col. i on the right.

Ο ṬA ΠΡ̣ ωϹ Τω̣ €Υθ̣ . Ξ̣Α̣ ⸎ΠΛΑ ΠΟΤ ΤΑΝΑ ΤΟΙ . €ΚΤ̣€̣ ΠΡωΤ ΤΥΙΔ̣Α ΟΥΚ€Δ̣ ΠΡΙΝϹ ΚΑΙΘΥ ΝΥΝΔ̣ . ΚΑΤΤΟ ΑΓΝΑΚ ΠΑΡθ̣ . ΑΜ̣ΦΙϹ̣Ọ Μ€ΤΡΟΛ †† . .

. . . . . Ν€ϹθΑΙ . . . . . Α̣Ϲ̣€ΠΑΚ̣ΡΑϹ ΝΧ . ΟΝΑ̣Δ̣€̣ΠΟΛΛΑ ω̣ΝΑΠ€̣́Χθ̣ΗΝ . . Υ̤Ν̤ΑΤΟ̣ . . ΙΝΑϹΓ̤ΑΡ Α̤Λ̤ . ϹΤΑϹ̣[ . . . . . ]ΟΝ . . . ΟΙϹΑ[ . . . . . ]ω . €ΟΡΤ̣ . . Π̤Ο̤ΗϹΑΝ Λ̣ΟΙϹΑ€θΛΟΙ . Αψ€ΡΟΝΔ€̣ ΟΝΓΑΡ€̣ΥΡ̣ . Π̣€Δ€ϹΛθΗΝ Π . ΙΔΑ̣ . . Π̣ΟΗΜ€Ν ΛΟϹ Υ . ΑΙΚωΝ

Ο[ Ṭα[ Πρ̣[ Ωσ[ Τω̣ Εὖ θ[ Ἐξ̣ ἀ̣[ Πλα[ Ποτ[ Τὰν ἀ[ τ' οἰ . [ Ἐκτ̣ε̣[ Πρῶτ[α Τυίδ̣' ἀ[ Οὐκ ἐδ̣[ Πρὶν σ[ὲ Καὶ θυ[ Νῦν δ̣ὲ [ Κὰτ τὸ [ Ἄγνα κ[αὶ Παρθ̣έ[ν Ἀμ̣φὶ σ̣ὸ̣[ν Μέτρ' ὀλ[ολύσδην †† . .

] . . . γένεσθαι ] . . . . . α̣ς̣ ἐπ' ἄκ̣ρας ]ν χίον'· ἀ̣ δ̣ὲ̣ πόλλα ]ω̣ν̣ ἀπέ̣χθ̣ην ] . δύ̤ν̤ατ'· ὄ̣ττινας γ̤ὰρ μ]ά̤λ̤ιστα σ̣[ίνν]ον‹τ'› ] . . . οισ' ἀ[γέσθ]ω ] . ἐόρτ̣α ] . π̤ό̤ησαν- λ̣οις ἀέθλοι[ς ] . ἄψερον δὲ̣ ὄδ]ον γὰρ ε̣ὔρ̣η[ν ] π̣εδέλθην ] παῖδα̣ ] . . π̣όημεν ὄχ]λος γ]υναίκων

Naturally, our poem starts at the coronis (i.e. the ⸎), that is l. 8 of the left column and l. 9 of the right one, and the lines before that are from another poem, which will appear on the blog between Dec 19 and sometime in January. I know some read more than me there, but really, the image doesn't allow even disappeared letters to come from the papyrus, so either some other fragment was reattached there, or I don't know how I could read e.g. προσ at l. 3 on the left when the papyrus margin barely contains the rho. [I assume this only refers to ll. 3-5 of fr. 2a col. ii, which were integrated with P.Oxy. 1231 fr. 36 found below (or above?).] The line in cruces on the left is:

I have seen it read as π̣α̣σ̣, but where are those letters exactly? I can kind of see the pi, the alpha may be weird-shaped but OK, but the sigma? Where? Given that my first reading was APH, the WTF when I saw ΠΑϹ was unimaginable. Since I cannot see where the letters start or stop, I just put cruces, it's a locus desperatus to me. I may put 5 vestiges though, two of which hanging over the missing bit, that is, two of which only half present. Below I retrace ll. 6-7 and 11 on the right, hoping to thus justify the critical notation:

Adding this P.GC. to the combination above

Let us now add this info to the above combined text, keeping the colors of before, and coloring the new info cyan:

Πλάσιον δή μ[] . . . οισ' ἀ[γέσθ]ω
Πότνι' Ἦρα, σὰ χ[αρίεσ]σ' ἐόρτ̣α
Τὰν ἀράταν Ἀτρ[εΐδα]ι π̤ό̤ησαν-
τ' οἰ βασίληες
Ἐκτελέσσαντες μ̣[εγά]λ̣οις ἀέθλοι[ς
Πρῶτα μὲν πὲρ Ἴ̤[λιο]ν, ἄψερον δὲ̣
Τυίδ' ἀπορμάθεν̣[τες· ὄδ]ον γὰρ ε̣ὔρ̣η[ν
Οὐκ ἐδύναντο
Πρὶν σὲ καὶ Δί' ἀντ̤[ίαον] π̣εδέλθην
Καὶ Θυώνας ἰμμ[ερόεντα] παῖδα̣
Νῦν δὲ κ[] . . π̣όημεν
Κὰτ τὸ πά[λαιον
Ἄγνα καὶ κα̣[λ      ὄχ]λος
Παρθε[νων      γ]υναίκων
Ἀμ̣φὶ σ̣ὸ[ν
Μέτρ' ὀλ[ολύσδην]
[]
. . α̣ν̣ιλ[
Ἔμμεν[
ρ̣απι[

P.Oxy. 7: line endings for Cypris and Nereids

An image of this papyrus can be found on Grenfell and Hunt Oxyrhynchus volume 1. Here is English Wikipedia's B&W version of what could be the same image:



And this is my transcription:

]Ν̤ΗΡ̈ΗΪΔ€̣ . . . ΛΑΒΗ[ ] . ΝΗΤΟΝΔ[ . ]Τ€ΤÝΙ̤ΔΙ̤Κ€̤Ϲ̣ΘΑ̣[ ] . ΘΥΜωΚ€̤Θ€ΛΗΓ€Ν€ϹΘΑΙ.1 ] . €ϹΘΗΝ· ]Ο̣ϹΘ’ÁΜΒΡΟΤ€ΠΑΝ . A̤ΛΥϹ[ ]ΙϜΟΙϹΙΧÁΡΑΝΓ€Ν€́ϹΘΑΙ ]Χ̤ΘΡΟΙϹΙ ̄ €ΝΟΙΤΟΔ’ΑΜΜΙ ]ΗΔ€ΙϹ ] . ΗΤΑΝ̣Δ€[ . ]€ΛΟΙΠÓΗ̤ϹΘΑ[ ]ΤΙΜΑϹ[ . . . ]ΙΑΝΔ€ΛÝΓΡΑΝ· ]Ο̣ΤΟΙϹΙ[ . . . ]ΟΙΘΑΧ€̣ΥωΝ ] . ΝΑ ]Μ̤€ΙϹΑΪω[ . ]ṬΟΚ€ΓΧΡω ] . €̤ΠΑΓ̣[ . . . ]ĀΙΠΟΛΙΤ̤ΑΝ ] . Λ€̣ỊΠ̤[ . . . ]ΝΗΚ€Δ’ΑΥΤ’ΟΥ ]ΚΡω̣ ]ΟΝΑỊΚ[ . . . . . . . . . ]Ϲ̤Ι ] . . Ν·ϹΥ[ . ]Κ̤ΥΠ̤[ . . ]Ι . [ . . . ]ΝΑ ] . €Ν̣ΑΚΑΚ̣ÄΝ[ ]ΙϹ̤ 1This is an actual dot in the papyrus: a period? A half-vanished comma?

] Ν̤ηρήϊδε̣ς, ἀβλάβη[ν κασί]γνητον δ[ό]τε τυί̤δ' ἴ̤κε̤σ̣θα[ι ] . θύμῳ κε̤ θέλη γένεσθαι τε]λέσθην· πρ]ό̣σθ' ἄμβροτε πάντα̤ λῦσ[αι ]ι ϝοῖσι χάραν γένεσθαι ἔ]χ̤θροισι, γ̣ένοιτο δ' ἄμμι μ]ήδεις κασιγ]νήταν̣ δὲ [θ]έλοι πόη̤σθα[ι ] τίμας, [ὀν]ίαν δὲ λύγραν ] ὄτ̣οισι [πάρ]οιθ' ἀχε̣ύων ] . να ]μ̤' εἰσαΐω[ν], τ̣ὸ κ' ἐγ χρῷ ] . ἐ̤παγ[ . . . ]ᾳ πολίτ̤αν ] . λε̣ı̣π̤[. ν ὄ]νηκε δ' αὖτ' οὐ μά]κρω ]ον, αἴ κ[ε . . . . . . . . ]σ̤ι ] . . ν, σὺ [δὲ,] Κ̤ύπ̤[ρ]ι, . [ . . . ]να ] . εν̣α κακ̣αν[[ ]ις̤

A couple reading notes:

• Concerning the γένοιτο, I formerly read that as ΓΕΝΟΓΙΟ or ΓΕΝΟΓΤΟ, but looking better I see the top stroke of the tau is ever so slightly detached from the preceding iota, making it impossible to read that as a gamma and something else:
  
  
  
• OK guys; I have no idea why everyone reads l. 18 differently in the papyrus, but I simply fail to see λυγ[ ]ρε[ ]να; I mean, look at this possible completion:
  
  
  is it really not plausible? OK, it’s a bit wide on ταἶνα and a bit compressed on δὲ, but why is the latter universally accepted as a completion and the former simply left unconsidered? I mean, the only source with something similar is TCPOS, in the others I’m lucky to find κυπρ instead of λυγρ!
• I seem to remember seeing the following line as θεμ[έν]α κακαν[, and I initially saw a possible theta there too, but squeezing in two letters between the (supposed) half-mu and the alpha is just not possible, so the next best thing is to assume a rather wide nu with a small gap between it and the alpha; we cannot adjust the positioning of the scrap with the alpha as we please since it's joined to the left part further down.
• Then we have the λειπ line. Image:
  
  
  
  The actual reason for the uncertainty is P.GC., but if you look at the beginning of this line, you can see how the ειπ could actually be ω, hence the uncertainty in the above transcription; and P.GC. will settle it to the omega.
• To finish off, start of l. 18. We see basically a right semicircle (not complete at the top), then an apparent top semicircle, and then, at the beginning of the lower portion of the line, a backslash that ends into a right semicircle with a dot above. I'd say the latter is a pretty convincing nu missing a left portion. As for the rest, the top semicircle could be the top of an omicron, and the other thing could be a slanted stroke for a nu, so ]Ν̣Ο̣Ν is an option, and ἴαι]νον is allowed. The top semicircle could as well be the top of a vertical stroke plus some quirks, and the other arc could belong to a rho, making ]Ρ̣Η̣Ν, that is φέ]ρην, another option. A certain Italian Anthology gives γνώσετ' ἂψ οἶον. You could read the first trace as the remnant of the iota, the top semicircle as an omicron, and get Ο]Ι̣Ο̣Ν = ἂψ ο]ἶον, so even that could be.

P.Oxy. 2289 fr. 6: middle portions of 3 lines of the above?

The overlap is small, but the fragment is also small, which gets the overlap to be almost all of the fragment. Do we accept joining it to the above? I guess we should. But of course, back in the days, when I translated Sappho, I had no idea this existed, so I have to reject it for the text in the post, and since the completion including P.GC. destroys mine completely, I can't use that either, so I am left with no completions if I accept this, and so I'll keep the pre-P.GC. version of the poem without this, and only include it in the post-P.GC. version in the all-Sappho posts. Anyway, image:

Transcription:

Ο̣Υ €Ο[.] ΚÝΠ̤

ο̣ὐ εο[.] Κύπ̤[ρι

After seeing another uncertain omicron turn to a sigma, I tried fitting that l. 1 into the text of the other papyrus, and the most convincing placement was as the συ in ἐσύνηκε. It's still not very convincing to me, but here is how it could be retraced:



I will not bother tampering with the post for this poem about this, but the editions should already have the συ out of the brackets. Why Lobel-Page and followers didn't notice this is beyond me. Did they seriously think of putting this as the οὐ in οὐκ ἄλλως or the one in δ' αὖτ' οὐ-? This clearly forces either ἄλλως or λειπ' [ἐ]σύνηκε. Note that the above retracing uses the completion from the collage with GC. The word ἐσύνηκε is attested as used by Alcaeus, so we can safely supplement it here, since it's Aeolic, and fits the context.

UPDATE
Lobel-Page reads that first line as ΝΗ̣, thus placing it two letters further right than I ended up doing. Looking at it, a certain nu is definitely not warranted, since the left leg is gone and the rest looks like an omicron more than part of a nu, so I'll have it double-underdot. As for the eta, my upsilon was probably unwarranted, since, looking at the upsilon in the last line, the difference is striking. Still, as half an eta, it is weird. Hence the underdot.
This makes the pertinence of this scrap to this poem unescapable, so I will probably add a middle version to the editions, with the not-too-good completion I came up with in the post for this poem, and also probably think harder, because that completion puts the εο way too far right. I will definitely at least have to fix the bracket notation in the editions, and re-put that συ in the square brackets.

Cypris and Nereids without P.GC.

Combining this with the above yields:

] Ν̤ηρήϊδε̣ς, ἀβλάβη[ν
κασί]γνητον δ[ό]τε τυί̤δ' ἴ̤κε̤σ̣θα[ι
] . θύμῳ κε̤ θέλη γένεσθαι
τε]λέσθην·

πρ]ό̣σθ' ἄμβροτε πάντα̤ λῦσ[αι
]ι ϝοῖσι χάραν γένεσθαι
ἔ]χ̤θροισι, γ̣ένοιτο δ' ἄμμι
μ]ήδεις

κασιγ]νήταν̣ δὲ [θ]έλοι πόη̤σθα[ι
] τίμας, [ὀν]ίαν δὲ λύγραν
] ὄτ̣οισι [πάρ]οιθ' ἀχε̣ύων
] . να

]μ̤' εἰσαΐω[ν], τ̣ὸ κ' ἐγ χρῷ
] . ἐ̤παγ[ . . . ]ᾳ πολίτ̤αν
] . λε̣ı̣π̤' [ἐσύ]νηκε δ' αὖτ' οὐ
μά]κρω

]ον αἰ κ[λ]έο[ς . . . . . ]σ̤ι
] . . ν, σὺ [δὲ,] Κύπ̤[ρ]ι, . [ . . . ]ΝΑ
] . εν̣α κακ̣αν[[
]ις̤

P.GC. inv. 105, fr. 3 col. ii and fr. 5: line beginnings for most lines of the above poem

This was found in 2014, way after I did my translations, and destroys my completion. It's a sibling of the other P.GC. used for Goddess Hera, and it's in the same image as that, so no image again. Here is my reading:

fr. 3 col. ii . ω̣ϹΠ̤ε̣ †† [ ΚΑΙΓΑΡ€̣ . [ Δ̣€Υ€Τω̣ . [ ΝΥΞΤ€ΚΑΙ̤ [ † . . ϹωΤ̣Ϲ̣ . . Ο̣Α̣ . . . . . †[ . . ΑϹω . . [ . . Α . . Α . . [ ΜΥΡΙ'̣Α̣ . . . [ Π̤Ι̣ΝΑ . . . . . [ ΠΟΤΝΙΑΙΝ [ ΤΟΝ̣Κ̣ΑϹΙΓΝ̣[ Κω . . ΙϜω [ ΚΗΝΟΤ€Λ€[ ΟϹϹΑΔ̤€Π . [ ΚΑΙΦΙΛ̣ΟΙϹΙ [ ΚωΝΙΑΝ€[ Μ̣Η̣ΔΑΜΑΜ[ ΤΑ̣Ν̣ΚΑϹΙΓΝ[ . . ]ϹΔΟΝΟ̣[ . . . . ]O[ fr. 5 . . ]Μ̣[ ω̣Ϲ̣ΠΟΤΟΥ̣[ Δ€ΝΔΙΑ . [ ΚΑΙΤΙΜ . [ Ν̣ωϹ€̣ . [ Ο̣ΥΚ̤Ο̤Ν̤ . [

fr. 3 col. ii . ὠ̣ς π̤ε̣†† [ Καὶ γὰρ ε̣ . [ Δ̣εύετ' ὤ̣ρ[α Νύξ τε καὶ̤ [ ††[ . . ασω . . [ . . α . . α . . [ Μύρι' ἄ̣στρ[α Π̤ι̣να . . . . . [ Πότνιαι Ν[ Τὸν̣ κ̣ασίγν[ητον Κὤττι ϝῷ [ Κῆνο τελέ[σθην Ὄσσα δ̤ὲ π . [ Καὶ φίλ̣οισι [ Κὠνίαν ἐ[ Μ̣ή̣δαμα μ[ Τὰ̣ν̣ κασιγν[ήταν Μέ]σδονο[ς . . . . ]ο[ fr. 5 . . ]μ̣[ Ὤ̣ς̣ ποτ' οὐ̣[κ δεν διὰ . [ Καί τι μ . [ Γ]ν̣ώσε̣τ[αι Ο̣ὐκ ὄ̤ν̤ε[κτον

• First line:
  
  
  
  What I see here is:
  
  • Room for a letter;
  • A possible half-omega where the left half is so faint it could well not even be there, making this letter an omicron;
  • A undoubtable sigma;
  • A very faint letter which might be a tau, an ypsilon, or, with a little stretch, a pi that has lost its right half to fading;
  • An epsilon;
  • What looks like a rho with a downward hook on the left; the reading ὤσπερ is indeed a strong temptation, but this papyrus usually doesn't hook rhos, so the hook is an objection to this being a rho; it could be a beta with the vertical stroke on the left having faded completely, leaving a kinda 3-shaped thing; so maybe ὤστε β or ὤστ' ἔβ[η; the cruces are because this kinda 3ish thingy could well be the left vertical stroke of an eta, if it were not curved; the rest of this non-eta is also weirdly curved and may be an alpha; ὤστ' ἔβα? Cruces because I cannot split the letters with certainty.
• The epsilon in l. 2 is very faint, but it cannot be anything else if it's not an illusion;
• The tau in l. 3 has a slanted horizontal stroke, but you can see said stroke does not continue into the vertical one as ypsilons usually do; the delta is uncertain because I cannot decide whether it's a delta or an alpha, though of course δεύετ' is a known word while *ἀεύετ' isn't AFAIK.
• In l. 4, for the xi I just trust the book the image is from, as that letter seems a line to me, and I don't think I've ever seen a xi in a papyrus; the iota at the end is so incredibly faint it could be illusory;
• Let me use this item for an image of the rest of the lines outside the Nereids poem, then I'll comment in the following items:
  
  
• L. 5 gets its gruces because I can't tell if it's one line or two; if it were two, the first one would look like interline glosses or corrections reading like the ruby text, and the second one would be the one under the ruby; this seems to be totally ignored by the book the image is from, which is why I fear it's not two lines, since the only thing the book has is the apparent ruby text; whatever the case, mostly nothing meaningful can be read out of this;
• In l. 8, MYPI is clear, the following A is barely visible after what looks like an apostrophe, and then I see room for 3 letters, with a possible sigma in second place, but it could also be the top of a tau;
• I clipped it off, but that line has something to the left of it which looks like a M;
• In l. 9, we have the faintest pi ever, then an iota which may be mistaken for the right half of said pi, then a N with a closed bottom that looks like an omicron, then an alpha, then room for 5 letters, the first two and last one of which have left some traces; I marked 5 vestiges because there is no hole in the papyrus, so using square brackets would seem inappropriate;
• In l. 11, the nu has lost its right vertical stroke, the kappa has lost its top slanting stroke and looks like a nu when joined with the alpha;
• The taus in κὤττι are barely visible, actually one has lost its vetical stroke and the other one its horizontal stroke;
• In ὄσσα δὲ, the delta merges with a fold(?) in the papyrus; the final vestige of l. 14 is just a high dot, definitely compatible with a rho; I initially missed it;
• In Καὶ φίλοισι, the lambda looks like a chi;
• The μήδαμα is badly messed up, hard to retrace, especially MH;
• In Τὰν κασιγν[ήταν, the tau is clear enough, the alpha partially visible, the nu looks like mu;
• I post the image of fr. 5 in the hope of justifying my critical notation without notes:
  
  

Adding P.GC. to Cypris and Nereids (which is now "divine Nereids")

So let us combine these lines with the text from the other papyri, keeping the above colors intact and coloring this papyrus yellow:

Πότνιαι Νηρήϊδε̣ς, ἀβλάβη[ν
Τὸν̣ κ̣ασίγνητον δ[ό]τε τυί̤δ' ἴ̤κε̤σ̣θα[ι
Κὤττι ϝῷ θύμῳ κε̤ θέλη γένεσθαι
Κῆνο τελέσθην·
Ὄσσα δ̤ὲ πρό̣σθ' ἄμβροτε πάντα̤ λῦσ[αι
Καὶ φίλ̣οισι ϝοῖσι χάραν γένεσθαι
Κὠνίαν ἔχ̤θροισι, γ̣ένοιτο δ' ἄμμι
Μ̣ή̣δαμα μήδεις
Τὰ̣ν̣ κασιγνήταν̣ δὲ [θ]έλοι πόη̤σθα[ι
Μέ]σδονο[ς] τίμας, [ὀν]ίαν δὲ λύγραν
Ἐκλύ]ο[ιτ'] ὄτ̣οισι [πάρ]οιθ' ἀχε̣ύων
] . να
]μ̤' εἰσαΐω[ν], τ̣ὸ κ' ἐγ χρῷ
. . ]μ̣[] . ἐ̤παγ[ . . . ]ᾳ πολίτ̤αν
Ὤ̣ς̣ ποτ' οὐ̣[κ ἄ]λλω̣ς̣, [ἐσύ]νηκε δ' αὖτ' οὐ
δεν διὰ μ[ά]κρω
Καί τι μᾶ[λλ]ον αἰ κ[λ]έο[ς . . . . . ]σ̤ι
Γ]ν̣ώσε̣τ[αι . .] . . ν, σὺ [δὲ,] Κύπ̤[ρ]ι σ[έμ]να,
Ο̣ὐκ ὄ̤ν̤ε[κτον καττθε]μέν̣α κάκ̣αν [ὔβ-
ριν,      ]ι

P. GC. inv. 105, fr. 2a col. i: line endings for "Dearest Offspring"

This is again a P.GC., so I cannot insert the image. Here is my transcription:

]ΟΗϹΑΙ ]€Θ̣ΟΙϹΑ ]Ο̣ΝΑΝ[] ]ϹΑ ]ΟΚΗωΝ ]Α̣ΓΑΥΤΑΝ ]Ν̣ΟΗΜΜΑ ]. ΝΟ̣ΗϹΗ ]ΝΕΜΝΑΙ ]ΒΑΜΑ ]Ν̣Π̣ΡΟϹωΠω̣ ]ΟΙϹΙ ]€Ν€̣ϹΘΑΙ ]ΧΗΝΔΑΡΑϹΘΑΙ ]Ṃ . ỴṬΑΙ

π]όησαι σκ]έθ̣οισα τ]ὸ̣ν ἄν[δρα] ]σα τ]οκήων ]α̣γ' αὔταν ] ν̣όημμα ]. νο̣ήση ὀ]νέμναι- ] βᾶμα ]ν̣ π̣ροσώπω̣ ]οισι γ]ένε̣σθαι έ]χην δ' ἄρασθαι ἐ]μ̣αύ̣τ̣ᾳ

Reading notes:

• The theta in l. 2 is doubtful because the horizontal stroke in the middle is half-gone, making it look like an omicron;
• The omicron in the following line is right on the edge;
• The nu in l. 9 is half-cut but clear enough;
• In l. 10, there is a vestige on the edge, then a half-faded but clear enough nu, then a blot which looks like an omicron, then HCH very clear;
• The epsilon in l. 11 does look a bit like a theta, but if you look closely it doesn't close on the right, making it impossible for it to be one;
• In l. 14, the nu is half-cut by the papyrus's edge, the pi has lost the top stroke and could be two iotas, the final omega looks like a weird omicron plus an iota because the connection (i.e. the second bottom arc) has faded away;
• The last line is terribly mutilated; the mu has lost the bottom of its vertical legs, but is clear enough; the alpha has lost everything save for its tip, hence just a vestige; the upsilon has left what looks like a curvy pit, which would probably be its top; the tau is barely visible and the top horizontal stroke is weirdly slanted; the AI is clear.

I would now combine this with the P.Oxy.s, but I haven't been able to clear up the mystery of 2166(a) or XXI add. p. something, so I'll wait before combining.

The rest of P.GC. inv. 105: fr. 1, fr. 3 coll. i-ii, fr. 4

P.GC., so no image. Off to the transcriptions:

Α̤Τ̣€̣Ρ̣€ΟΡΤΑΝ . ẠΝωΡ̣Α̤ỊΤ€Λ . ΑΜ€ΡωΝ€Μ Ν̤ḤΝ̤Θ̣Α̣Ϲ . Μ ỌNΑKΟYCΑỊ Υ̣Ν'ΟΥΤΟϹ . ΗΝÝΝ̣Α̣ΒΛ̣Α̣ ẠCΔΙΔωΝΤΑ ΟΗϹ€Ν· . . . . . €̣ΠΙϹ̣Ϲ̣ Ν̣ΥϹΤΟΝ€̣ Ν . Τ€Λ€ϹΘ̣Η Δ̣€̣ΓωΠΑΜΠ . ΑΝ . ΓΛωϹϹΑ . ΤΑ̣ΠΥΓΝώ . ΡΟΦ€́ΛΛΗϹ̣ €ΡωΝ€ . . ΠΑ̣Ν̣ fr. 3 col. i ΑΡ̣ . ϹΤ̣Α . . . . . ϹΑΝ . . Θ̣ΥΜΟΝ Γ̤ΑΡ . ΡΜω̣Ϲ · Α̤ΙΝ ω fr. 3 col. ii Ϲ̤ωϹΤ€Ρ . [ ΚΑΙΓΑΡΕ . [ Δ€Υ€ΤωΡ̣[ ΝΥξΤ€ΚΑΙ̣[ . . οωΟϹΤ . [ . . ΑϹω . . [ . . Α . . Α . . [ ΜΟΜΥΡΙÁϹ̣ . [ +. .+ΤΟ̣ΝΑ . . . [ fr. 4 ΠΑϹ̤ €ΑΜ€ / €ΜΑ€ ΠΟΛ ΟΤΤΙ

μ]ᾶ̤τ̣ε̣ρ̣ ἐόρταν ] . αν̣ ὦρ̣α̤ι̣ τελ[ ] . αμέρων ἔμ[ ]ν̤η̣ν̤ θ̣ᾶ̣ς . μ[ ]ο̣ν ἄκουσαι̣ ]υ̣ν' οὖτος .[ ] ἢ νῦν̣ ἀ̣βλ̣α̣[ ]α̣ς δίδων τα[ π]όησεν· . . . . . ]ἐ̣πισ̣σ̣[ ]ν̣υστονε̣[ ]ν . τελέσθ̣η[ν ] δ̣' ἔ̣γω πάμπ[αν ] . αν . γλῶσσα [ ] . τα̣πυγνώ[ ] . ρ ὀφέλλης̣ ] ἔρων ε . . [ ]πα̣ν̣[ fr. 3 col. i ἄρ̣ιστ̣α . . . . . σαν . . θ̣ῦμον γ̤ὰρ . ρμω̣ς · α̤ιν ω fr. 3 col. ii σ̤ῶς τε ρ . [ καὶ γὰρ ε . [ δεύετ' ὤρ̣[ νύξ τε καὶ̣ [ . . οστ . [ . . ασω . . [ . . α . . α . . [ {μο} μύρ̣ι' ἄ̤σ̤ . [ +αὖ+το̣ν ἀ . . . [ fr. 4 πᾶς̤ ἔλλαβ[ πολ[ ὄττι

Reading notes for fr. 1:

• L. 4 looks like NMN more than NHN and the theta is missing the middle stroke and looks like an O and the alpha resembles a lambda, hence the uncertainties. No idea why the vestige after the sigma is given as an epsilon.
• L. 5 may seem like MAKOYCAN because the nu's diagonal stroke doesn't touch the bottom of the right vertical one and the I seems to continue into a N.
• ABLA in l. 7 and ΔΙΔωΝΤΑ in l. 8 have uncertain alphas and lambdas because they can't be told apart easily.
• In l. 11 the epsilon is missing the middle stroke hence the uncertainty, then I see the pi, a tall vertical line going 2 lines down, the I, then I assume that would be taken as a kappa by Obbink for some reason, but there is no vertical stroke to the left of the sigma-like part which could be the two slanted strokes, unless we take the I for that vertical stroke and then the I would go into that tall line which seems bollocks to me, then what Obbink treats as a vestige definitely looks like a C.
• In l. 12 the N has left only its right vertical stroke making it look like an I, the epsilon has a faint curvy part and no middle horizontal stroke.
• In l. 13 the theta is easy to take for a C because the top hole is shrunk to almost nothing and the bottom hole is left unclosed to the right.
• The delta in l. 14 is barely visible and the epsilon looks like a sigmα but ΔϹΓ is not a thing in Greek and €Γω is just too compelling.
• In l. 15 again an alpha looking like a lambda.
• No idea why the last line is not transcribed by Obbink. Maybe the ΠΑΝ is treated as from another fragment?

Reading notes for fr. 3 col. i:

• ΘΥΜΟΝ has its theta hardly visible and with no middle stroke.
• Two lines below, the ΡωϹ is a messy scribble:
  • The rho is kind of visible, it connects to whatever follows.
  • The omega has its right "valley" closing into a wannabe omicron – indeed I read MO C at first.
  • However, I cannot have that, because there is no valid explanation for what comes before: the left leg of the mu would bed downward as if it was superimposed with the right slanting stroke of an upsilon which then terminated in the vertical bottom, then the upsilon would have the left slanting stroke visible, and there would be a vestige at the start, but the superposition is just too suspicious.

Reading notes for fr. 3 col. ii:

• In l. 1, the sigma is barely visible and even a potential papyrus quirk, hence the double underdot. I may be Edmondsing my way to those letters ("many words even outside the brackets are very doubtfully legible", as he comments to his version of LP 92).
• In l. 3, the omega seems to me to be the only option that matches the traces, hence it being upgraded from Obbink's underdot to my certainty.
• In l. 5, the tau seems hardly doubtable, given its very non-faint state. The following letter seems to be an epsilon, but could also be an omicron that doesn't close to the right.
• In l. 6, the alpha is weirdly shaped and almost twice as tall as the following letters. One might read an upsilon after the omega, but also a gamma, or even a tau if they squint hard enough.
• In l. 7 I straight up refused to try reading that mess of traces, and just reported the sufficiently clear alphas.
• The MO in l. 8 is definitely spurious as it's contra metrum; perhaps it's a papyrus quirk. It is weird that the line starts slightly indented. No matter how hard I squint, I just cannot see two traces after the sigma. There is a strange space between alpha and sigma, with a super-faint tau in it, which is probably a papyrus quirk.
• In l. 9, I can't see how Obbink read π̣ι̣. There is indeed a tau or half-pi after a little indent (hence my +..+), but I can't see the other half, and the iota isn't there. One could try to see a pi in that +..+ and a nu over the omicron, whose bottom could be a papyrus quirk, but between the two there is a tau, with a very clear horizontal top. Looking at it again, what I read nu could actually be a mu, and then the alpha is clear enough, and then traces. So we could be looking at τομα, though the mu would run into the alpha this way. I maintain my original reading.

Reading notes for fr. 4:

• The sigma in ΠΑϹ is barely visible;
• The two middle letters of l. 2 are joined and the line could even be €ΛΛΑΒ, hence the right column;
• The TT in OTTI could be a pi, though it seems too wide to be one.

Actually, let me give you the picture of fr. 4:

• The first comment is that that double-underdot sigma is so barely visible that I will make it a vestige.
• L. 2 is hard. The first epsilon is, I'd say, certain. The rest is confusing. We have that first ^, then what is either a mu or two lambdas, then what is the left half of either an epsilon or a beta, though an epsilon is more likely because the top and bottom are expanding upward and downward respectively whereas a beta would have them vertically still until they stooped in the opposite directions. Could be €ΑΜ€, €ΜΑ€, €ΑΜΒ, €ΜΑΒ, €ΑΛΛ€, €ΑΛΛΒ, €ΛΛΑ€, €ΛΛΑΒ. Anything else? I find it exceptionally unconvincing, but someone proposed it's €ΙΜ€. The only way I could see that is if the left side of that ^ was a connection to the epsilon, and thre right side was a slanted iota. I'm keeping my original idea, but taking this up for the combination (see at end of post).
• Looking closely at l. 4, the first tau's top is slightly slanted up-left, which suggests it might actually be a flattened upsilon, so to speak, and the final letter is curved, thus suggesting ΟΥΤΟ.

So I revise my transcription:

ΠΑ . [ €Λ̣Λ̣Α̣Β̣[ ΠΟΛ[ ΟΥ̣ΤΟ[

πᾶς [ ἔλ̣λ̣α̣β̣[εν πολ[ οὖ̣το[

P.Oxy. 2289, frr. 4 & 10: more of the poem from P.GC. inv. 105 fr. 1 above, and an extra fragment the image sticks to it

As the title says, the extra fragment is just stuck to fr. 4 in the image:

And here is the transcription of both:

Α̣ΡΚΑΛ€ΙϹ̣ΙΤΑϹ€Λ̣ Τ̣ΑΝΟΥΚ€ΧΗ €Ρ€ÓΡΤΑΝ Μ̣ΑΝ[ . ]ΡΑΙΤ€Λ€ . ωΝ€́Μ ÃϹÁΜΚωϹΖω ΥϹΑΙ ΟϹΔ€ . . . fr. 10 ṂẠ ṆΟΝΘ Π .́ .

π]α̣ρκάλεισ̣ι τασελ̣[ ]τ̣αν οὐκ ἔχη [ ]ερ ἐόρταν ]μ̣αν [ὦ]ραι τέλε[σαι . ωνέμ ] ἆς ἄμ[μεςκὠς ζῶ[μεν υσαι ος δὲ . . . fr. 10 μ̣α̣ ν̣ονθ π .́ .

Reading notes:

• The first alpha in l. 1 is half torn away; the uncertain sigma closes to an apparent omicron; the final lambda looks like a 1;
• The starting tau in l. 2 is half torn away;
• The starting mu in l. 4 is half torn;
• The ending mu of l. 6 is half torn and could be a nu; no idea how Obbink gives a certain epsilon in the gloss in place of my kappa, the only way to see an epsilon is to imagine a reversed three shaped epsilon where the two curves were drawn along a slanted line.

Combining the above Oxyrhynchus fragment with the matching GC fragment

So let me combine P.GC. inv. 105 fr. 1 with P.Oxy. 2289 fr. 4, coloring the former red:

π]α̣ρκάλεισ̣ι τασελ̣[
]τ̣αν οὐκ ἔχη [
μ]ᾶ̤τερ ἐόρταν
]μ̣αν ὦραι τέλε[σαι
ἐ]παμέρων ἔμ
]ν̤η̣ν̤ θ̣ᾶς ἄμ[μες
]ο̣ν ἄκουσαι
υ̣ν' οὖτος δὲ
] ἢ νῦν̣ ἀ̣βλ̣α̣[
]α̣ς δίδων τα[
π]όησεν·
. . . . .
]ἐ̣πισ̣σ̣[
]ν̣υστονε̣[
]ν . τελέσθ̣η[ν
] δ̣' ἔ̣γω πάμπ[αν
] . αν . γλῶσσα [
] . τα̣πυγνώ[
] . ρ ὀφέλλης̣
] ἔρων ε . . [

P.Oxy. 1231, fr. 1, part of col. i: the biggest chunk of the pre-P.GC. text of Sappho 16

I report here the transcription, and some notes, found on P.Oxy. X, since P.Oxy. 1231 is not available in online images. Note that the last line is from col. ii and combines with PSI II 123.

[.]ΙΜ€ΝΙΠΠΗωΝϹΤΡΟΤΟΝΟΙΔ€Π€ϹΔωΝ ΟΙΔ€ΝΑωΝΦĂĨϹ€Π[.]ΓΑΝΜ€ΛΑΙ[.]ΑΝ [.]ΜΜ€ΝΑΙΚΑΛΛΙϹΤΟΝ·€ΓωΔ€ΚΗΝ'ÓΤ ΤωΤΙϹ€ΡΑΤΑΙ [..]Γ̣ΧΥΔ€́ΥΜΑΡ€ϹϹÝΝ€ΤΟΝΠΟΗϹΑΙ [.]Α̣ΝΤΙΤ[.]Υ̣Τ̣'ᾹΓΑΡΠÓΛΥΠ€ΡϹΚΟ̣Π̣€̣Ι̣[.]Α [...]Λ̣Ο̣Ϲ̣[...]ΡωΠωΝ€Λ€ΝΑ̣[..]ΝΑΝΔΡΑ [.........]Ι̣ϹΤΟΝ [.......]Ϲ̣€ΒΑϹΤΡΟΪΑ[..]Λ€ϹϹ̣[..] [......]Α̣Ị̃ΔΟϹΟΥΔ€Φ̣[.]Λ̣ωΝΤΟ[.]ΗωΝ [.......]€ΜΝĀ́ϹΘ'Α[...]ΠΑΡĀ́ . . Γ̣'ΑΥΤΑΝ [........]ϹΑΝ [.......]ΑΜΠΤΟΝΓΑΡ[ [....] . . . ΚΟΥΦωϹΤ[.......]ΟΗϹΗ̣ [..]Λ̣€̣ΝΥΝΑΝΑΚΤΟΡΙ[...]Ν̣ . ΜΝΑ         ΜΝΑΙ [..]Η̣ΑΠ̣€ΟΙϹΑϹ. [..]ϹΤ€ΒΟΛΛΟΙΜΑΝ€Ρ̣ᾹΤΟΝΤ€ΒΑΜΑ Κ̣ᾹΜÁΡΥΛΜΑΛΑΜΠΡΟΝΙΔΗΝΠΡΟϹωΠω Η̣ΤΑΛΥΔ̣ωΝΑΡΜΑΤΑΚΑΝΟΠΛΟΙϹΙ [.......]ÁΧ€ΝΤΑϹ [.......]Μ€ΝΟΥΔΥΝΑΤΟΝΓ€Ν€ϹΘΑΙ [.....]Α̣ΝΑΝΘΡωΠ[....]€Δ€ΧΗΝΔᾸΡᾹCΘΑΙ [Several lines lost] Τ€ΞΑ̣[

[Ο]ἰ μὲν ἰππήων στρότον οἰ δὲ πέσδων Οι δὲ νάων φαῖσ' ἐπ[ὶ] γᾶν μέλαι[ν]αν [Ἔ]μμεναι κάλλιστον· ἔγω δὲ κῆν' ὄτ- τῳ τις ἔραται. [Πά]γ̣χυ δ' εὔμαρες σύνετον πόησαι [Π]ά̣ντι τ[ο]ῦ̣τ̣'· ἀ γὰρ πόλυ περσκό̣π̣ε̣ι̣[σ]α [Κάλ]λος [ἀνθ]ρώπων Ἐλένα̣ [τὸ]ν ἄνδρα [....... αρ]ι̣στον [.......]σ̣εβας Τροΐα[..]λεσσ̣[..] [Κωὐδὲ π]α̣ῖ̣δος οὐδὲ φ̣[ί]λ̣ων το[κ]ήων [.......] ἐμνάσθ' ἀ[λλὰ] παράγαγ' αὔταν [........]σαν [.......]αμπτον γὰρ [ [....] . . . κούφως τ[.......]οήση̣ [Πῆ]λ̣ε̣ νῦν Ἀνακτορί[ας ὀ]ν̣εμναί- [σθ]η̣‹ν› ἀπ̣εοίσας. [Τᾶ]ς ‹κ›ε βολλοίμαν ἔρ̣ατόν τε βᾶμα Κ̣ἀμάρυ‹χ›μα λάμπρον ἴδην προσώπω Ἢ̣ τὰ Λύδ̣ων ἄρματα κἀν ὄπλοισι [......μ]άχεντας [.......]μεν οὐ δύνατον γένεσθαι [.....]α̣ν ἀνθρώπ[οις, π]εδέχην δ' ἄρασθαι [Several lines lost] τ' ἐξ ἀ[δοκήτω.

Notes

• The accusative γᾶν μέλαιναν at l. 2 is unexpected, and read thus on the basis of the same expression in the Hymn to Aphrodite; another option is to read a genitive plural μελαίναν and suppose an error for γᾶς or γᾷ;
• The reading of l. 6 is very uncertain, and indeed has given two very distinct options. At the end, σκ is followed by a rounded letter, ε, ο, possibly α; next to it, a rather high stroke turning over to the left, which would suit ρ or perhaps δ; the termination may be either οι[σ]α or ει[σ]α; so σκεδοῖ[σ]α is just as likely as σκοπεῖ[σ]α, and by that "stroke turning to the left" description I assume σκεθοῖ[σ]α is also just as likely;
• Near the beginning of l. 7, an interlineated α, δ, or λ is more likely than a grave accent; and below this are vestiges of what seem to have been round letters; the omission of one of the lambdas of κάλλος is a not unlikely error;
• παράγαγε seems to be the right word in l. 11, and γα is possible, though not suggested by the very small vestiges remaining from the tops of the letters;
• The supposition of a corruption in a mutilated papyrus (as in l. 16) is generally objectionable; moreover the π of ἀπεοίσας, though not impossible, is really more like ρ, i.e. ] παρεοίσας, not ]η ἀπεοίσας, is the more natural reading; but it seems difficult to adapt this to the preceding remains and the apparent sense; if, as would rather be gathered from the gist of the whole poem, Anactoria was absent, οὐ must preced παρεοίσας, and there might also be room in the lacuna for another letter, e.g. κοὐ or -σ'οὐ; in l. 15, ]νεμνα is suitable, but μεμνα is equally possible; of the ε there is hardly anything left; for the marginal μναι cfr. the spelling μέμναισθ', ὀμναῖσαι in the Berlin fragment (aka P.Berol. 9722 (fol. 2 to be precise), aka "I really wish I had died"); at the beginning of the line λ is hardly certain, and σ might well be substituted;
• The mark of length on ἔρατον may be due to confusion with ἀρατόν;
• The alpha in -μάχεντας at l. 20 is written through an epsilon;
• Last stanza restored by W-M as «Εὖ μὲν ἴδμεν οὐ δύνατον γένεσθαι / Τοῦτ' ἀν' ἀθρώπων, πεδέχην δ' ἄρασθαι / Ἔστι πὰρ θέων μακάρων ἔκοισαν / Τῶν παρεόντων», «Well we know that it is not possible for this to happen to men, but it is possible, for one willing, to pray to have a share in the present gods».

P.Oxy. 1231, fr. 36: more of the above poem

A small scrap doubtfully assigned to this poem by Lobel-Page which P.GC. basically confirms is to be assigned to the end of this poem.

ΠΡΟϹ[ ωϹΔ[ [..] . [

προσ[ ωσδ[ [..] . [

P.Oxy. 2166(a), fr. 2: two scraps to be combined with the above poem

Unfortunately, even P.Oxy. vol. XXI doesn't give a raw transcription of these. All I can see is «Two scraps have been attached to 1231 fr. 1 i so that ll. 19-24 (Σμ ᾱ 5, 7-12[, that is 7-12 of the above transcription]) are now to be read», and the combined text. I decided to try and guess what the papyrus might have looked like, and this is my attempt at recreating it:



Let's make a couple comments:

• The red lines show the borders of the scraps, while the margin of the big 1231 fr. 1 is black;
• The chopped-off kappa could be a chi or a mini-lambda too, hence the vestigial nature in the combined text;
• Same for the lambda-bottom; the interlineated lambda is split between fr. 1 and the scrap, being mostly on the scrap, so it will be assigned to the scrap;
• OC and P might be less or more cut, fact is the combined text gives them as vestiges so pretty little must have been there, but still enough for GH to venture a (probably pretty daring IMHO) guess;
• Analogous comment for the A of ELENA and the I in the following line, and the C formerly in CEBAC;
• The second scrap has a pretty weird shape; the leftward fingerlike protrusion is just for that apostrophe, which may have been missed by GH; the rightward fingerlike protrusion, instead, is necessary to explain (at least partially) the misterious transformation of letters in l. 9 which is the main reason for doing this; the waviness is probably because I spaced out the letters in l. 10 a little too much;
• The kappa and pi starting ll. 9 and 11 have been cut out to form similar arguments to the above, and the same goes for the tau of l. 8; only the grave accent survives from l. 12, and that is known from the combined text;
• Similar cutting arguments apply at the margin of fr. 1 in ll. 8-10, and whatever else is chopped off by the margin;
• The phi and lambda of ΦΙΛωΝ are mostly in fr. 1, hence the guess by GH, but now that they're complete, they're certain; still, assigned to fr. 1, and rendered as vestiges in the separate transcription of the second scrap;
• Now my crux: the λεσσ̣ to λεοι̣ transformation; I guess the first sigma (aka omicron) was certain to GH because they saw it detached from the rest, and then the second sigma (aka iota) was curved, so they thought maybe sigma, but it's too cut off to be sure; then the fingerlike protrusion appeared, and revealed a blank at the bottom, so the sigma was excluded; at this point, it was basically inevitable to posit that a vertical stroke had accidentally been curved there, and looked like a sigma before the protrusion, which ruled any round letters out; so the guess was made that this was the simplest letter with a vertical stroke: an iota; and looking at πλε . ι[, with the context they had, they said «this has to be an omicron», and made it certain, with a very daring and IMHO not justifiable move; so I'll make it uncertain.

That said, transcriptions. In the second scrap, I ignore the left fingerlike protrusion, and put the right one only in l. 9, where it gives the +.+.

] . Α .Λ [ ] . Ο . [ ] . ΑΛΛ[ ]ΚὼΥΔ[ ] . Α[ .] .̀ [

Κάλλ[λος Τὸν [ Καλλ[ Κωὐδ[ὲ Πά[μπαν .] .̀ [

]ΝΠ[. . .]+.+[ ] . Ι . [ ]ΛΛΑ[

]νπ[. . .]+.+[ ] . ι . [ ]λλα[

Combining the Oxyrhynchus sources for this poem

Below, I give the pre-Oxy.XXI text on the left and the post-Oxy.XXI text on the right. Oxy 1231 1 i is non-colored, Oxy 1231 36 is red, Oxy XXI 2166(a) is blue.

[Ο]ἰ μὲν ἰππήων στρότον οἰ δὲ πέσδων Οι δὲ νάων φαῖσ' ἐπ[ὶ] γᾶν μέλαι[ν]αν [Ἔ]μμεναι κάλλιστον· ἔγω δὲ κῆν' ὄτ- τῳ τις ἔραται. [Πά]γ̣χυ δ' εὔμαρες σύνετον πόησαι [Π]ά̣ντι τ[ο]ῦ̣τ̣'· ἀ γὰρ πόλυ περσκό̣π̣ε̣ι̣[σ]α [Κάλ]λος [ἀνθ]ρώπων Ἐλένα̣ [τὸ]ν ἄνδρα [Κρίννεν ἄρ]ι̣στον [Ὂς τὸ πᾶν] σ̣εβας Τροΐα[ς ὄ]λεσσ̣[εν] [Κωὐδὲ π]α̣ῖ̣δος οὐδὲ φ̣[ί]λ̣ων το[κ]ήων [.......] ἐμνάσθ' ἀ[λλὰ] παράγαγ' αὔταν [........]σαν [.......]αμπτον γὰρ [ [....] . . . κούφως τ[.......]οήση̣ [Πῆ]λ̣ε̣ νῦν Ἀνακτορί[ας ὀ]ν̣εμναί- [σθ]η̣‹ν› ἀπ̣εοίσας. [Τᾶ]ς ‹κ›ε βολλοίμαν ἔρ̣ατόν τε βᾶμα Κ̣ἀμάρυ‹χ›μα λάμπρον ἴδην προσώπω Ἢ̣ τὰ Λύδ̣ων ἄρματα κἀν ὄπλοισι [......μ]άχεντας [.......]μεν οὐ δύνατον γένεσθαι [.....]α̣ν ἀνθρώπ[οις, π]εδέχην δ' ἄρασθαι [Several lines lost] Πρὸς [ Ὠς δ[ [Several lines lost] τ' ἐξ ἀδοκή[τω.

[Ο]ἰ μὲν ἰππήων στρότον οἰ δὲ πέσδων Οι δὲ νάων φαῖσ' ἐπ[ὶ] γᾶν μέλαι[ν]αν [Ἔ]μμεναι κάλλιστον· ἔγω δὲ κῆν' ὄτ- τῳ τις ἔραται. [Πά]γ̣χυ δ' εὔμαρες σύνετον πόησαι [Π]ά̣ντι τ[ο]ῦ̣τ̣'· ἀ γὰρ πόλυ περσκέ̣θ̣ο̣ι̣[σ]α Κ̣άλλος [ἀνθ]ρώπων Ἐλένα̣ [τὸ]ν ἄνδρα Τὸν [ άρ]ι̣στον Καλλ[ίποι]σ̣' ἔβα' ς Τροΐαν πλέοι̣[σα] Κωὐδ[ὲ π]α̣ῖ̣δος οὐδὲ φίλων το[κ]ήων Πά[μπαν] ἐμνάσθ' ἀλλὰ παράγαγ' αὔταν [.] .̀ [......]σαν [.......]αμπτον γὰρ [ [....] . . . κούφως τ[.......]οήση̣ [Ὤς] μ̣ε̣ νῦν Ἀνακτορί[ας ὀ]ν̣εμναί- [σ' οὐ] π̣αρ̣εοίσας. [Τᾶ]ς ‹κ›ε βολλοίμαν ἔρ̣ατόν τε βᾶμα Κ̣ἀμάρυ‹χ›μα λάμπρον ἴδην προσώπω Ἢ̣ τὰ Λύδ̣ων ἄρματα κἀν ὄπλοισι [......μ]άχεντας [.......]μεν οὐ δύνατον γένεσθαι [.....]α̣ν ἀνθρώπ[οις, π]εδέχην δ' ἄρασθαι [Several lines lost] Πρὸς [ Ὠς δ[ [Several lines lost] τ' ἐξ ἀδοκή[τω.

Throwing P.GC. into the mix

Here is the text above, with the colors kept, and GC added in, colored yellow. Note that the first yellow bits are from fr. 2a col. i, while the ones at the end come from fr. 2a col. ii and fr. 2b col. i. For the last line, the contribution of PSI II 123 is pink. I saved brown and purple for two distinct Athenaeus quotations formerly inserted in another poem from Oxy 1231 which P.GC. forces to insert here. Contrary to usual practice, I will leave the part of the quotations found in P.GC. as they are found there, rather than making everything certain since it's in a quotation.

[Ο]ἰ μὲν ἰππήων στρότον οἰ δὲ πέσδων
Οι δὲ νάων φαῖσ' ἐπ[ὶ] γᾶν μέλαι[ν]αν
[Ἔ]μμεναι κάλλιστον· ἔγω δὲ κῆν' ὄτ-
τῳ τις ἔραται.

[Πά]γ̣χυ δ' εὔμαρες σύνετον πόησαι
[Π]ά̣ντι τ[ο]ῦ̣τ̣'· ἀ γὰρ πόλυ περσκέθοισα
Κ̣άλλος [ἀνθ]ρώπων Ἐλένα̣ [τ]ὸ̣ν ἄνδρα
Τὸν [ άρ]ι̣στον

Καλλ[ίποι]σ̣' ἔβα' ς Τροΐαν πλέοι̣σα
Κωὐδ[ὲ π]α̣ῖ̣δος οὐδὲ φίλων τοκήων
Πά[μπαν] ἐμνάσθ' ἀλλὰ παράγα̣̣γ' αὔταν
[.] .̀ [......]σαν

[.......]αμπτον γὰρ [ ] ν̣όημμα
[....] . . . κούφως τ[.....] . νοήση̣
[Ὤς] μ̣ε̣ νῦν Ἀνακτορί[ας ὀ]νεμναί-
[σ' οὐ] π̣αρ̣εοίσας.

[Τᾶ]ς ‹κ›ε βολλοίμαν ἔρ̣ατόν τε βᾶμα
Κ̣ἀμάρυ‹χ›μα λάμπρον ἴδην προσώπω
Ἢ̣ τὰ Λύδ̣ων ἄρματα κἀν ὄπλοισι
[......μ]άχεντας

[.......]μεν οὐ δύνατον γένεσθαι
[.....]α̣ν ἀνθρώπ[οις, π]εδέχην δ' ἄρασθαι
] ἔγω δ' ἔμ̣' αὔ̣τ̣ᾳ
Τοῦτο σύνοιδα.

[Possible lacuna of some whole stanzas.]

. . . γένεσθαι
Ο[      ] . . . βὰ̣ς̣ ἐπ' ἄκ̣ρας
Τ̣α[      ]ν χ[ί]ον'· ἀ̣ δ̣ὲ̣ πόλλα
Πρὸς

Ὠς δ[      ]ω̣ν ἀπέ̣χθ̣ην
Τω̣[      ] . δύ̤ν̤ατ', ὄ̣ττινας γ̤ὰρ
Εὖ θ̣έω, κῆνοί με μά̤λ̤ιστα σ̣ίννον-
τ' ἐξ ἀδοκή[τω.

P.Oxy. 1231, fr. 15: the so-called "Gongyla poem"

I copy both the transcription in P.Oxy. X and that in Lobel-Page, because there are wildly different versions of this poem with differences whose justifications essentially boils down to "lectio nonnullis locis difficillima" (reading is extremely hard in some places), as Lobel-Page puts it, and to a cascade of effects of inserting the name Ἄβανθι in l. 2 on the basis of the Alcaeus fragment from P.Oxy. 2299 fr. 10, which ends with ὄρχησθ' [ἐρό]εσσ' Ἄβανθι, as opposed to taking this poem as addressed to Gongyla as P.Oxy. X does. Left column is P.Oxy. X raw and P.Oxy. X with some restorations, right column is Lobel-Page stripped of restorations and diacritics and spaces and then Lobel-Page as it is in the book. Each column has the reading notes of the source its transcriptions are taken from.

Grenfell-Hunt [.] . Ν̣ . [.........]€ΛΟΜΑΙϹ̣[ [..] . ΓΥΛ̣Α̣ . [....]ΝΘΙΛΑΒΟΙϹΑΜ̣Α̣ . [ [...]Κ̣ΤΙΝΑΝ̣Ϲ̣€ΔΗỸΤ€ΠΟΘΟϹΤ̣ . [ ΑΜΦΙΠΟΤΑ̣ΤΑΙ ΤΑΝΚΑΛΑΝ·ĀΓΑΡΚΑΤÁΓ̣ω̣Γ̣ΙϹΑΥΤΑ̣[ €ΠΤΟΑΙϹ'ΙΔΟ̣Ι̣ϹΑΝ·€ΓωΔ€ΧΑΙΡω· ΚΑΙΓΑΡÁΥΤ̣Α̣ΔH́Τ[..]€Μ€ΜΦ̣[ .]Υ̣Π̣ΡΟΓ€Ν[ .]Α̣ϹĂΡĀΑΜΑ̣[ ΤΟΥΤΟΤῶ[ .]ΟΛΛΟΜΑ̣[ [.] . ν̣ . [........ κ]έλομαί σ̣[ [Γο]γγύλ̣α̣ . [....]νθι λάβοισα μ̣α̣ . [ [γλα]κ̣τίναν̣ σ̣ε δηῦτε πόθος τ̣ . [ ἀμφιπότα̣ται τὰν κάλαν· ἀ γὰρ κατάγ̣ω̣γ̣ις αὔτα̣[ ἐπτόαισ' ἴδο̣ι̣σαν, ἔγω δὲ χαίρω· καὶ γὰρ αὔτ̣α̣ δὴ τ[όδ]ε μέμφ̣[εταί σοι Κ]υ̣π̣ρογέν[ηα· τ]ᾶ̣ς ἄραμα̣[ι τοπυτο τὦ[πος β]όλλομα̣[ι

Lobel-Page .] . €̣ . [....] . [....]€ΛΟΜΑΙϹ . [ ..] . ΓΥΛΑ . [...]Α̣ΝΘΙΛΑΒΟΙϹΑ . Α . [ ..]ΚΤΙΝΑϹ̣Ϲ€ΔΗΥΤ€ΠΟΘΟϹΤ̣ . [ ΑΜΦΙΠΟΤΑΤΑΙ ΤΑΝΚΑΛΑΝ·ĀΓΑΡΚΑΤÁΓωΓΙϹÁΥΤ̣Α[ €ΠΤΟΑΙϹ'ΙΔΟΙϹΑΝ·€ΓωΔ€ΧΑΙΡω· ΚΑΙΓΑΡÁΥΤ̣ΑΔH́ΠΟ̣[.]€Μ€ΜΦ[ .]ΥΠΡΟΓ€Ν[ ω̣ϹĂΡĀΜΑ̣[ ΤΟΥΤΟΤῶ[ .]ΟΛΛΟΜΑ̣[ .] . ε̣ . [....] . [... κ]έλομαί σ . [ Γο]γγύλα . [...]α̣νθι λάβοισα . α . [ ..]κτιν, ἆς̣ σε δηὖτε πόθος τ̣ . [ ἀμφιπόταται τὰν κάλαν· ἀ γὰρ κατάγωγις αὔτ̣α[ ἐπτόαισ' ἴδοισαν, ἔγω δὲ χαίρω, καὶ γὰρ αὔτ̣α δήπο̣[τ'] ἐμέμφ[ετ Κ]υπρογέν[ηα ὠ̣ς ἄραμα̣[ι τοῦτο τῶ[ β]όλλομα̣[ι

No uncertainty notes in GH, only interpretation notes.



As for LP:

• «.]δε .[ vel .]βε . [      ]τ[ vel ]γ[      κ] suppl[evit] e[adem] p[apyrus]      σα[, σε[, σο[ possis», that is, in l.1, the uncertain epsilon may be preceded by delta or beta, the unreadable letter between lacunas may be a tau or a gamma, the kappa of κέλομαι is a restoration from P.Oxy. X, σα σε σο are the options at the end of the line; which means Campbell's version, the vulgate version and Edmonds' version are all possible -- well yeah, that epsilon destroys the beginning of Edmonds, but the rest is possible;
• «Γο]γγυλα e[adem] p[apyrus], sed ]ργυλα vel ]'γυλα scriptum potius credideris; tum ν, π, β, ι, alia, ante lac[unam] legere possis      fort[asse] Ἄβ]ανθι, cf. Alc. M 10 (b) i 8      fin[e] inter α et α litt[eram] μ vel ν legas», that is, in l.2, Γο]γγύλα isn't the most natural reading, but the suggestion of P.Oxy. X; before the following lacuna, ν, π, β, ι, or others may be read; supplementing Ἄβ]ανθι is possible in light of the Alcaues fragment I mentioned above (which is M 10 (b) aka number 261 in Lobel-Page (and nr. 261 in Campbell), and has 8 lines); the doubtful letter between alphas near the end may be μ (Oxy version) or ν (Campbell); so again, all versions I know of are possible;
• «κτιν· fort[asse] πᾶ]κτιν (quod Castiglione post λάβοισα, v. 10, proposuit). incert[um] utrum ας̣ an α[.] scribendum sit», that is, maybe πᾶκτιν is to be read (as proposed by Castiglione), and it is uncertain whether one should write ἆς̣ or ἆ[ς]; we are on l. 3;
• «[…] γα[ pro τα[ legi potest», that is, the doubtful tau near the end of l. 5 may be a gamma;

Well I thought all problems were solved, but the αν̣ => ας̣ one stays (though with the α[.] option one can write α[ν] and be content), and that certain tau in δήτ[..] became a certain pi in δήπο[.] in l. 8? That is quite a mystery… That goes to show how different people may be certain about different readings and different uncertainties because reading papyri is hard AF. Not having an image, that is all I can say. The only other source I can find that may belong here is an addendum in P.Oxy. XXI, but that is only in the same column as this and fr. 12, and seems to fit in neither fragment, and indeed is not included by Lobel-Page or by Campbell. It is also interesting to note how vestiges magically appear in Lobel-Page that P.Oxy. X didn't have. I sure wish I had an image to clear these mysteries up… One thing to be noted is that, though Voigt's edition doesn't give much more about this, it did help me figure out that the extra ].[ in l. 1 from Lobel-Page is from l. 9 of fr. 12, which P.Oxy. X reads ]τ̣[, so if we accept to join the two fragments, that letter is no mystery (though there is no mention of a possible alternate reading of it in P.Oxy. X…). But that is it. The only way to clear this up is really to have an image, but the only P.Oxy. 1231 fragment available online is fr. 56, found on Wikimedia Commons and in the Wikipedia page about "Papyrus Oxyrhynchus 1231", which I will transcribe soon. So this mystery will have to stay, since the Bodleian library decided not to upload images of P.Oxy. 1231 which they have.
And then, on Feb 4 2018, I googled P.Oxy.2166(a), and Google went "SURPRISE MOTHERF'ER", landing me on an image of a number of P.Oxy. 1231 (and 2166(a)) fragments, one of which was this one. So here is the image of fragments 12 and 15 and a couple intruders I don't feel like erasing from the image. The top-right thing is a part of fr. 14, for which see below. The other one is P.Oxy. 2166(a) fr. 1.

Transcription:

fr. 12 ]ΒΛ̣ . . [ ]€ΡΓΟΝ· . +.+Λ̣ÁΤ€[ ]ΝΡ€ΘΟϹΔΟΚΙ̣[ ]ΗϹΘΑΙ ]Ν̣ΑΥĀ́ΔΗΝΧ . [ ]€Μ̣H́Χ€ΙΜω̣ . [ ]Τ̣ΟΙϹΑΝΑΛΓ€Α . [ ]Δ€ ] . [ fr. 15 ] . . . [ . . . . . . . . . . ]€ΛΟΜΑΙϹ[ ] . ΓΥΛΑ̣ . [ . . . . ]ΝΘΙΛΑΒΟΙϹ̣Α̣ . . [ ]Κ̣ΤΙΝ . [.] Ϲ̣€ΔĤΥΤ€ΠΟΘΟϹ̣ . [ ΑΜΦΙΠΟΤ̣ΑΤΑΙ ΤΑΝΚΑΛΑΝ·ÁΓΑΡΚΑΤÁΓω̤ΓΙϹÁΥ[.] . [ €ΠΤΟΑΙϹ'ΙΔ . [.]ϹΑΝ·€ΓωΔ€Χ̣Α̣ΙΡω· ΚΑΙΓΑΡÁΥ . Α̣Δ̣H́ . [. .]€Μ€Μ . [ .]ΥΠΡΟΓ€́Ν[ .] . ϹΑΡ̣ĀΜ . [ . ΟΥΤΟΤῶ[ ]ΟΛΛΟΜ . [ P.Oxy. 2166(a) fr. 1 (bottom-right) ] . ΟΙϹΑ̣[ ]ΥΡΙΝΝΟΙ ] . ΑΥΤΑ̣Ν ]Ϲ'€ΟΙϹΑΙ̣Ϲ̣ ]ΛΟΙϹΑ

fr. 12 ]βλ̣α . [ ]ἔργον· ἀ+λ+λ̣ά τε [ ] ρέθος δοκί̣[μ ]ησθαι ]ν̣ αὐάδην χ . [ δ]ὲ μ̣ὴ χείμω̣ν[ ]τ̣οισαν ἄλγεα . [ ]δε ] . [ fr. 15 ] . . . [ . . . . . . . . . κ]έλομαι σ[ Γο]γγύλα̣ . [ . . . . ]νθι λάβοισ̣α̣ . . [ ]κ̣τιν . [.] σ̣ε δηὖτε πόθος̣ . [ ἀμφιπότ̣αται τὰν κάλαν· ἀ γὰρ κατάγω̤γις αὔ[τ]α[ ἐπτόαις' ἴδο[ι]σαν· ἔγω δὲ χ̣α̣ίρω· καὶ γὰρ αὔτα̣ δ̣η . [. .]εμεμφ[ Κ]υπρογέν[ηα· .] . ς ἄρ̣αμα[ι τοῦτο τὦ[πος β]όλλομα[ι P.Oxy. 2166(a) fr. 1 (bottom-right) ] . οισε̣[ Γ]ύριννοι ] . αυτα̣ν ]σ' ἐοίσαι̣ς̣ ἔ]λοισα

We thus proceed to reading notes answering the questions from above. But first, in l. 3 of the 2166(a) fragment, the Ạ is uncertain because I took it for an O originally, and then wanted to align with the Lobel-Page text.

• Joining the fragments still seems a bit precarious, as no argument of sense continuity is allowed by fr. 12, and no physical joining is present; fiber matching may help, but the picture doesn't highlight it;
• On l. 1 of fr. 12, the first letter may look like an alpha at first sight, and I fell into that trap, but then I realized it looks nothing like the other alphas in P.Oxy. 1231 fragments, where the horizontal/slanted stroke joins the right leg waaaay higher; a delta is also to be excluded, because the deltas here are consistently isosceles triangles with a certain base length, and making this into an isosceles triangle would make the base waaaay too long; all other letters save for beta are easily ruled out, hence this is a beta; the following one could be an alpha, but it seems less likely, hence its status as underdot lambda; then we have what is probably an alpha's belly, but I went prudent and made it a vestige; and finally, there is a low dot that GH had missed too, but which is part of a descender, and hence to be placed in l. 1 as a vestige;
• On l. 2, the lacuna is not a hole, and in fact there are traces of ink which seem to spell "OY." there; with some positing of papyrus quirks, we can reduce the traces to a dot, which forms a colon with the high dot, and a possible half-alpha;
• O l. 5, «the nu and upsilon are certain, and in the middle is possibly a trace of ink, though I'd see more of an apostrophe than an alpha» is what I thought at first, but that nu consisted of a vertical leg, a slanted stroke which I now realize is probably the effect of a minor papyrus quirk that produces a joining illusion, and part of a pretty solid alpha, and there is even half of a slanted stroke starting from the bottom of that vertical leg, so an uncertain nu and a certain alpha are there; the vestige is a vertical line, which could be an iota, but also several different letters, as it's right on the margin;
• On l. 6, you may be surprised to see a certain epsilon, but a horizontal line halfway up the line can only be the horizontal stroke of an epsilon; the last three letters at the end are an ugly scribble which may in fact allow a reading MM; the mid-strokes of omegas don't reach lettertops, so none of those can be; if you zoom in enough, the first vertical line is detached from the rest, so we have an iota, then a mu, and then a trace which AFAICT has no signs of rising back up to lettertops, hence the uncertain omega status; another vestige is then found on the margin, which I have no clue why LP treats as a certain nu, it's literally just a vertical line;
• On l. 7, the tau has left only its top stroke, and part of it, so it surprises me that GH reads tau and not gamma as happened in fr. 51 (see below);
• After l. 8, there is a horizontal stroke left from a l. 9 which is otherwise totally lost; this gives the tau in Lobel-Page and Campbell, but it may well be any of the following: Γ, Ζ, Ξ, Π, Τ;
• The beginning of l. 1 in fr. 15 is terribly mutilated: only two strokes are visible, one horizontal, one slanted (but very little), and a little dot right on the corner; here are two retracings:
  
                  
  on the left, we take the slanting stroke as the one of a very short nu, and the horiontal stroke as a letter connection of a lost letter to an upsilon; this is quite unlikely IMHO, but not impossible; the more likely thing is on the right, where the slanting stroke is the bottom of an epsilon, and the horizontal one the bottom of a delta, though it could be a beta; [ΝΥΝ] Δ €[Γω is possible, saving one of my completions;
• The famous extra vestige is bollocks: there is NO INK after that sigma;
• I agree with LP that the ink trace on l. 2 suggests an apostrophe more than a gamma; the alpha is uncertain 'cause that bottom may be a chi, and then there is a vertical stroke which could be any of ΒΓΗΙΚΜΝΠΡ; in ΛΑΒΟΙϹΑ, the sigma and alpha have only left their tops, hence te uncertainties; then there is the top of a two-legged letter with no horizontal strokes, hence mu, nu, or eta, and a blot which could be any of ΑΔ€ΘΛΜΟ, though the pointy nature of that blot would naturally suggest ΑΔΛ;
• The kappa in l. 3 is held uncertain since the visible ink suggests chi as an alternative; the alpha is rendered a vestige because the bulk of it has vanished; I can see why LP was in doubt between ΑϹ and Α[.], since the alpha could include both traces or just one, but reading a sigma out of that is precarious given that there is a whole area where the ink faded out and that trace could be part of the slanting stroke of a nu; the sigma is only half-visible and could be an epsilon; same goes for the sigma near line end, and the last letter has only left a vertical stroke which allows for many options, including tau and gamma;
• The omega in l. 5 has left only its bottom, with a bit too many traces, hence the double underdot; the following gamma is certain because we can see the horizontal stroke and the bottom of the vertical one, and the horizontal one doesn't extend to the left of that bottom, thus excluding tau and leaving only gamma; after AY there is a hole and then a trace, very likely the belly of an alpha;
• In l. 6, the iota after the apostrophe looks doubled; the omicron following the delta has left two little traces, and the iota is totally gone, unless that potential trace is taken as a sign of that iota; XA have left only their tops, hence the underdots;
• In l. 7, I was probably feeling brave on the gamma, which is so messed up that I'm tempted to change it to uncertain; after AY there is a hole, then a horizontal stroke allowing tau or gamma, then the top of a possible alpha, the top of a possible delta, an H with acute accent, a tau or pi (the right vertical leg is partly out of the papyrus and partly on that blue spot, which renders it impossible to see if there is ink there or not), the uncertain omicron is justified by what seems to be two minimal traces which make it shorter than surrounding letters and require the following letter to be extra wide since the hole seems to be 2 letters wide, and the phi at the end has left only a dot;
• In l. 9, we see a dot and something on top, which rather suggests a messed-up partial omega, though the top thing could be some kind of damage and then we would have an alpha; at the end, the supposed alpha only left a caret-like thing which I leave as a vestige;
• In l. 1 of the 2166(a) fragment, the vestige is again caret-like, possibly the bottom of a lambda or alpha; there is a dot on the right of the hole, but I only just noticed it and putting a vestige in doesn't seem worthwhile; the last letter looks like an epsilon at first sight, but is slanted, while the other epsilons in P.Oxy. 1231/2166(a) are straight (meaning the curved arcs are left-pointing semicircles, not half-ellipses pointing up-left like this one is), and the slant would be consistent with an alpha;
• In l. 4, there is a dot on the margin which I only just noticed; the IC at the end are connected, but IC seems more likely than N, given the curve of what would be the right leg of the N; you can even almost see a horizontal stroke in the middle of the sigma, but AIE is not metrical here, so we will dismiss that as a papyrus quirk.

P.Oxy. 1231, fr. 56: "Let's sing for the couple!"

Well, surprise! A P.Oxy. 1231 fragment with online image! Here it is:



And here is my transcription:

Ν . . . [...] . ΠΑΡΘ€ΝΟΙΔ ΠΑΝΝΥΧΙϹΔΟ . ϹΑΝΑ€ΙΔΟΙ . ΦΑϹΙΟΚΟΛΠω ΑΛΛ€Γ€ΡΘ . ϹΤ€ΙΧΕϹÒΙϹ H́Π€ΡΟϹϹΟΝẠ ΥΠΝΟΝ[.]ΔωṂ Μ€ΛωN Α ΧΗΗΗ Δ̣Ο̣

Ν . . . [...] . Πάρθενοι δ[ὲ παννυχισδο .[ σὰν ἀειδοι .[ φας ἰοκόλπω ἀλλ' ἐγέρθε[ις στεῖχε σοὶς [ ἤπερ ὄσσον ἀ̣ [ ὔπνον [ἴ]δωμ̣[εν Μ€ΛωN Α ΧΗΗΗ Δ̣Ο̣

P.Oxy. 2166(a), fr. 6a: more to the above

I can only guess at the raw transcription of this by subtraction of the above from the combined text given in the addenda of P.Oxy. XXI.

]Α̣ . [ ]ΝΦ[ ]Ι̣Ϲ,ΗΪΘ[ ]ΥΜΑΛΙ . [ ]ΑΛΙΓΥΦ . [ ]€Ν

]α̣ . [ ]νφ[ ]ι̣ς, ἠϊθ[ε ] ὐμάλικ[ ] ἀ λιγύφω[νος ὄρνις ]εν

Comparing with the image of the combined fragments (which is together with that of Gongyla and many more at a link given above), it turns out that I got almost everything right. Two comments:

• The first line of this scrap actually reads ". .", or three dots, the leftmost of which seems somewhat round, hence suggesting the belly of an alpha, the middle one would be the right end of the alpha, and the rightmost one, right on the margin, would be the bottom of the iota;
• The article-alpha in the second-last line was completed by the second scrap, which gave the right bottom and made it certain; given that the bulk of the alpha was in the 1231 fragment, I'll color it black in the combined text;
• the upsilon looks more like a comma, so I'll underdot it; the omega is half-chopped off, hence the vestigial nature;
• the scrap completed the right vertical leg of the mu and gave the bulk of the epsilon, so the mu will be black and the epsilon blue.

Combining sources for "Let's sing for the couple!"

And the combined text, with the 2166(a) scrap colored blue, is:

Πάρθενοι δ[ὲ
παννυχίσδοι[σ]αι [
σὰν ἀειδοιεν φ[ιλότατα καὶ νύμ-]
φας ἰοκόλπω
ἀλλ' ἐγέρθεις, ἠιθ[έοις
στεῖχε σοὶς ὐμάλικ[ας
ἤπερ ὄσσον ἀ λιγύ̣φω[νος ὄρνις]
ὔπνον [ἴ]δωμ̣εν

P.Oxy. 1231 frr. 50-54 and P.Oxy. 2166(a) fr. 5 (two scraps): a poem on nostalgia for youth

Here is the image, again excerpted from this lot, on the left as is, on the right with the joining lines between the various fragments highlighted (the red line is where a tad of guesswork was involved) and the 1231 fragments numbered (the 2166(a) ones wouldn't allow fitting the numbers, and only have a single shared number anyway):

And here are my transcriptions:

fr. 50 ] . ΚΑΙΓΑΡ . [ ]Ι̣Ν̣ . ϹΜ€ . [ ]Ζ̣ΑΛ€ΞΑΙΚ[ ]ΔΡΑΧΑΡΙϹϹ . [ ] . €ΙΧΟΜ€ΝΓΑ . [ ]ΙϹΥΤΟΘΥΤ·ΑΛΛ[ ] . [.]€ΝΟΙϹÁΠ[ ]Ν€̣ΧΟΙ€Ν̣[ fr. 51 ] . ωΔ€Κ[ ] ]ΟΝ·€ΥΔ . [ ]ΑΧΙϹΤ̣Α[ ]€ΜΗΠ€·Θ€ΟΙ[ ] ] . ΑΝ€ΙϹΟΛ̣[ ]ĂΙΚ . [ fr. 52 ]Λ̣ΑΙΓ[ ]ΝΟϹ[ ] ]ΠΟΤ[ ]Τ̣ΑΥΤΑ[ fr. 53 ]Ο̣ΔΟ . . [ ]ΝΘΡω̣[ fr. 54 ]ϹΙ· ]ΗϹ . [ ]Π€Ϲ[ fr. 5a (leftmost 2166(a) scrap) ] . ΜΜ[ ]Α̣Ι ] . Ο€Ϲ[ ]Ο[ fr. 5b (rightmost 2166(a) scrap) ] . ΠΑΪϹ[ ]ΘΑΓΙΤ[ ]ΑΠΥ . [ ] ]ΓΑΜ[ ]ΤΤΙΤ[

fr. 50 ] . καὶ γὰρ . [ ]ι̣ν̣ . σμε . [ ] ζ̣άλεξαι κ[ ἄ]δρα χάρισσα[ι σ]τείχομεν γὰρ [ κα]ὶ σὺ τοῦτ'· ἀλλ[ πα]ρ[θ]ένοις ἀπ[ κε]ν ἔ̣̣χοιεν̣ fr. 51 ] τῶδε κ[ ] ]ον· εὖ δὲ [ τ]άχιστ̣α [ π]εμπε· θέοι [ ] μέ]γαν εἰς Ὄλ̣[υμπον ]αικ . [ fr. 52 ]λ̣αιγ[ ]νος[ ] ] ποτ[ ] τ̣αῦτα [ fr. 53 ] ὄ̣δος . [ ἀ]νθρω̣[π fr. 54 ]σι· ]ησ . [ ]πεσ[ fr. 5a (leftmost 2166(a) scrap) ] . μμ[ ]α̣ι ] . οεσ[ ]ο[ fr. 5b (rightmost 2166(a) scrap) ] . πάϊς [ ]θ' ἄγι τ[ ] ἀπὺ . [ ] ] γάμ[ον ὄ]ττι τ[

Reading notes.

• In l. 1 of fr. 50 we have a trace on the very edge, a mere dot, and another dot on the far right, before a discolored area;
• L. 2 features a vertical line which could be the bottom of an iota tau or gamma (or maybe other letters), a pseudo-nu traced by the margin and a vertical line which I give as Ṇ because there also appears to be a trace in the corner that goes back up, but I will turn to a Ḳ in the combined text, then a dot right above a small vertical line and a horizontal line sightly to the right which appears to be trying to end between dot and vertical line, which suggests either eta or epsilon, the former strengthened by this apparent dot above the end of the horizontal line, but this dot seems to continue into a curved line going down and strengthening the eta, and since I am sure you are as confused as me I guess you get why I chose the reading "." for this :);
• That same line ends with a rising stroke which could be either half a lambda or a quarter of a mu;
• The following line starts with kinda like an angle ∠, probably the bottom of a Z, the top horizontal stroke does seem to have left a trace, but I keep GH's uncertainty because of the clear mutilation of this letter;
• The following line starts with what is definitely half a delta to me and ends with a little curved stroke, which to my eyes only the 5a scrap reveals as the belly of an alpha and not e.g. a sigma
• L. 5 starts with a horizontal trace crashing into the epsilon, then features a mu whose left half is barely visible due to a hole, and ends with two letters: an alpha that is recognizable though chopped, and a vertical line which could be many things, though a little ending tilt does suggest a rho;
• L. 6 starts with the top of a vertical line which leaves no doubt about it being the top of an iota, then the tay of TOYT is right next to a hole but it cannot be a pi because the vertical leg we see is practically halfway through the horizontal one, making it impossible to get a decent pi out of it;
• That trace in l. 7 cannot be assigned to any letter IMHO, it is too trace-y;
• L. 8 starts with a dot taken as a trace, then a doubtless nu, then very faint ink suggesting a doubtful epsilon, then XOI€ without the horizontal stroke of the epsilon (or almost), then the nu is certain but partially faint (I mean, the right leg is visible, and the rest is terribly faint);
• L. 1 of fr. 51 starts with a horizontal trace I do not want to assign to a letter in the transcription, but which could be a tau or a gamma;
• The trace at the end of l. 3 does suggest an epsilon, but a sigma looks equally likely, and both are weird in that the curve is sharpened to an angle;
• The traces at the start of l. 5 indicate an undoubtful though very mutilated epsilon; the ruby denotes an eta corrected into a pi;
• L. 7 starts with a horiontal stroke which again could be a tau or gamma, hence the "." reading; the final lambda is right next to the margin, hence it could be a half mu, hence the uncertainty;
• The alpha to me is doubtless, especially given the breve above it, which makes it alpha omicron or upsilon, and since the traces don't allow any of the three but alpha…;
• The lambda in fr. 52 is very mutilated and looks like a caret, and is therefore uncertain, and also because GH have it uncertain;
• L. 1 of fr. 53 features this apparent vertical trace which suggested an iota to GH, but which has a connected horizontal trace leading to the margin above and below, making it a mutilated omicron to my eyes; after the ΔΟ, a hole with a dot next to it, and some trace afterwards;
• The omega in l. 2 of fr. 53 is over-half-chopped, hence its uncertainty;
• L. 2 of fr. 54 ends with this rather slanted stroke that probably suggested the tau to GH, but seems IMHO much too slanted for that, and indeed it joins to the traces in 5b to form an upsilon which will be marked uncertain and as belinging to 5b;
• L. 1 of fr. 5a starts with a low stroke going towards bottom-right, which combined with the trace over the first half of the first mu suggests Á;
• This fragment has, in l. 3, a caret completing the alpha, and a circle which, in isolation, seems to be a certain omicron, but when joined to the vertical trace of fr. 50 reveals its nature of rho-top, which is rendered certain and will be marked as belonging to 5a;
• Between the theta and alpha at the start of l. 2 of fr. 5b there is a trace which seems like a comma or high dot, though it seems out of place in either case.

Combining sources for the above poem

So finally we combine the text, marking fr. 50 as black, fr. 51 as blue, fr. 52 as red, fr. 53 as brown, fr. 54 as purple, fr. 5a as cyan, and fr. 5b as pink (what a lot of colors :) ). I omit the terribly holey first stanza.

ἦσθ]α καὶ γὰρ δὴ σὺ̣ πάϊς ποτ' [ἄβρα
ἐξ]ί̣κ̣ης μέλπεσθ' ἄγι ταῦτα [πάντα
σοὶ] ζ̣ἄλεξαι κἄ̣μμ' ἀπὺ τῶδε κ[ῆρος
ἄ]δρα χάρισσαι
σ]τείχομεν γὰρ ἐς γάμον· εὖ δέ [γ' οἶσθα
κα]ὶ σὺ τοῦτ'· ἀλλ' ὄττι τάχιστ̣α [
πα]ρ[θ]ένοις ἄπ[π]εμπε, θέοι [
κε]ν ἔ̣̣χοιεν̣
οὐ γάρ ἐστ'] ὄ̣δος μ[έ]γαν εἰς Ὄλ̣[υμπον
ἀ]νθρω[π      ]αικ . [

P.Oxy. 1231 fr. 10: a poem on old age

] . €ΠÁ . ΟΛΗ . [ ] . ΝΔÓΛΟ·ΦΥΝ[....] . [ ]ΤΡΟΜ€ΡΟΙCΠ . [...]ΛΛΑ ] ]ΧΡΟΑΓΗΡACHΔΗ ]ΝΑΜΦΙBACK€Ι ] . € . ΑΤΑΙΔΙωΚωΝ ]C̣I ]TACΑΓÁYAC ] . Α·ΛΑΒΟΙCA ] . €ICΟΝΑΜΜΙ ] ] . ωN;Μ̣Α̣ΛΙCTA ]ẠCΠ[..]ΝΑΤΑΙ̣

] . ἐπάβολ' ἦσ[κε ] . ν δολόφυν [....] . [ ]τρομέροις π . [.. ἄ]λλα ] ] χρόα γῆρας ἤδη ]ν ἀμφιβάσκει ] πέταται διώκων ]σ̣ι ] τᾶς ἀγαύας ] . α· λάβοισα δ' ἀδύφωον πᾶκτιν] ἄεισον ἄμμι τὰν ἰόκολπον.] ἔ]ρων μά̣λ̣ιστα ] ἆ̣ς π[λά]ναται̣

Reading notes:

• The high dot in l. 2 matches the dot on the top right of the omicron which precedes it; I will ignore it because δολο is not a valid ending for a Greek word AFAIK;
• The vestige in l. 2 is a vertical stroke with a slight hook to the left, suggesting a rho;
• Ll. 5-6 have weird extra signs on the right, probably either corrections or starts of lines from another column; I cannot make sense of those, hence I didn't transcribe them;
• L. 6 starts with the tops of three letters, one looking like a horizontal stroke with two small downward appendixes at the endpoints, perhaps a pi, then there is an undoubtful epsilon though most of it is barely visible, then two little horizontal strokes, probably a tau's horizontal stroke with the middle faded out;
• In the last two lines, the rho only left a sort of apostrophe, whereas the slanting stroke with ink at the bottom is very likely an alpha, uncertain because of mutilation;
• I am actually unsure whether l. 8 ends with what I transcribed or not; option two could allow for ἀλγε]σί[δωρον, though that would require a fair amount of layout fiddling since other lies are totally lost and should be wider.

P.Oxy. 1231 fr. 13 + P.Oxy. 2166(a) fr. 7a: (potentially) another poem on old age

This is two fragments joined. I will refrain from highlighting the one very small joint line, and just give the image of the fragments joied.



And here comes the transcription of both.

fr. 13 ] . Α̤ΝÁΓ . [ ]€ΜΝᾹϹ€ϹΘ'Α̤[ ]ΜΜ€Ϲ€ΝΝ€Ο̣[ ] . ΟΗΜ̣Μ€Ν. ]€ΝΓΑΡΚΑΙΚΑ[ ]ΜΕΝ·ΠΟΛΙ[ ]Ο[.]€ΙΑΙϹΔ[ fr. 7a ].[ ]ΑΙΓ̣ΑΡΑ[ ΤΑΥ̣Τ' . . [ ΠΟ̣Λ̣Λ̣Α̤[ . Δ . Η .[ ]ΜΜ€̤[ ] . . [.] . . [

fr. 13 ] . α̤ναγ . [ ]εμνάσεσθ' ἄ̤[ϊ ἄ]μμες ἐν νεό̣[τατι ε]πόημ̣μεν. μ]ὲν γὰρ γαὶ κά[λα ]μεν· πόλι[ ] ὀ[ξ]είαις δ[ fr. 7a ].[ κ]αὶ γ̣ὰρ ἄ[ ταῦ̣τ' . . [ πό̣λ̣λ̣α̤ [ . δ . η .[ ἄ]μμε̤[ ] . . [.] . . [

Reading notes:

• In l. 1 of fr. 13, the first letter is merely a short vertical stroke ending at the edge, and then we have what strongly suggests itself as the bottom of an alpha; the final letter may well be the right vertical leg of a Π, of which the Γ would constitute the rest;
• In l. 2, the N is missing its middle, but is nonetheless certain; the final letter is a low stroke rounded to the left, which could be an alpha, but also an omicron;
• The epsilon in l. 3 has lost its middle stroke, hence the doubtfulness; the final omicron is a high left-bellied stroke, too high for an alpha and too tight for a sigma or epsilon, hence the single underdot;
• The pi in l. 4 has only left its right leg, the left one being in the other fragment, and the horizontal top in the hole; the first mu has lost its left half to a hole;
• The epsilon in l. 7 is half-cut-off but still certain;
• The vestige in l. 1 of fr. 7a is just a dot that managed to survive the mutilation;
• The gamma in l. 2 has lost its top and could therefore also be a tau; the alpha is half-faded-out but still pretty much certain; the other alpha is half-cut away (the top is in the hole) but still certain;
• The right half and most of the vertical leg of the upsilon in l. 4 are faded away, hence the underdot; don't be fooled into connecting that little dot belonging to the alpha to the left side of the upsilon's top; after the apostrophes, a dot (top of an epsilon?) and a right-bellied stroke which was meant to be straight and the left leg of a pi;
• The pi in l. 4 is certain, the other letters are barely visible, and another one is faded away entirely;
• In l. 5, we have a trace near the edge, which Voigt suggests may be a zeta (or was it Lobel-Page?), then what is commonly given as a vestige but I wager is a pretty undoubtable delta, a low near-horizontal stroke, and a rather visible eta (albeit with partly faded right half), then some faint trace;
• The epsilon in l. 6 is almost entirely faded away;
• No point describing the unintelligible traces in the first part of l. 7; the second part seems to have a T and then aother trace, a dot in a little finger of papyrus.

Combination of the above

Combined text (7a in blue):

] . α̤ναγ . [
] . [    ]εμνάσεσθ' ἄ̤[ϊ
κ]αὶ γ̣ὰρ ἄμμες ἐν νεό̣[τατι
ταῦ̣τ' ἐπόημ̣μεν.
πό̣λ̣λ̣α̤ [μ]ὲν γὰρ γαὶ κά[λα
. δ . η . [    ]μεν· πόλι[
]μμε̤ [    ] ὀ[ξ]είαις δ[
] . . [.] . .

On 4/5 at 3:35, I came up with a completion possibility for the above, which I included as ROS 7. L. 2 of stanza 2 starts with a mess which I suppose is a verb, so I leave it thus, and have two completion options depending on how many syllables that verb has. The completion is, naturally, just a rehash of Edmonds.

Αἶσ' ἔγων ἔ]φα̤ν· ἄγα[ναι γύναικες, Ο]ἶ[α μ]εμνάσεσθ' ἄ̤[ϊ μέχρι γῆρας, κ]αὶ γ̣ὰρ ἄμμες ἐν νεό̣[τατι λάμπρᾳ Ταῦ̣τ' ἐπόημ̣μεν. Πό̣λ̣λ̣α̤ [μ]ὲν γὰρ γαὶ κά[λα κἄγν' ἐν αὔτᾳ . δ . η . [    ]μεν· πόλι[ν δ' ἀπυλιππανοίσαις ῎Υ]μμε̤ὀ[ξ]είαις δ[άκε μοι βλεποίσᾳ Θ]ῦμ[ο]ν ἄ[σαισι Ἴμμερος.

Αἶσ' ἔγων ἔ]φα̤ν· ἄγα[ναι γύναικες, Ο]ἶ[α μ]εμνάσεσθ' ἄ̤[ϊ μέχρι γῆρας, κ]αὶ γ̣ὰρ ἄμμες ἐν νεό̣[τατι λάμπρᾳ Ταῦ̣τ' ἐπόημ̣μεν. Πό̣λ̣λ̣α̤ [μ]ὲν γὰρ γαὶ κά[λα κἄγν' ἐν αὔτᾳ . δ . η . [    ]μεν· πόλι[ν αὖ λιποίσαις ῎Υ]μμε̤ ὀ[ξ]είαις δ[άκε μοι βλεποίσᾳ Θ]ῦμ[ο]ν ἄ[σαισι Ἴμμερος.

I will think of how to complete that line (i.e. try to look for a suitable verb), and if I succeed, I will add the result here. I had thought of Συμποήσαμεν or another suitable prefix, but the delta demolished it. Looking at the papyrus again on 5/5 around 16:39, I realize that delta is more of an omicron, which allows <>πονήσαμεν, or more precisely <>πονήσ̤[α]μεν, meaning completion 2 is the way.

P.Oxy. 1231 fr. 14: ode to a beautiful woman

] . ΡωΤΟϹΗ . Γ̣[ ] ]Τ̣Ι̣ΟΝ€ΙϹΙΔωϹ[ ] . ΡΜΙΟΝᾹΤ€Α . [ ]ΞΑΝΘΑΙΔ'€Λ€ΝΑΙϹ€ḮϹ[.]ΗΝ ]Κ€Ϲ ] . ΙϹΘΝΑΤΑΙϹ·ΤΟΔ€Δ'ÍϹ . [.] Ι̤ΤĀ́ΙϹÂΙ ] . ÁΙϹΑΝΚ€ . €ΤΑΝΜ€ΡÍΜΝᾹΝ ]ᾸΙϹ'ΑΝΤΙΔ[ . . . . . . ]ΘΟΙϹΔ€̤ ]    ΤΙ̤Η̤Ν̣ ]ΤΑϹΟΧΘΟΙϹ ]ᾸΙΝ ]ΝΥΧΙ . [.]Η̣Ν

] ἔρωτος ἤλγ̣[ει ] αν]τ̣ι̣ον εἰσίδω σ[ε ] Ἐρμιόνᾳ τεαύ[τα ] ξάνθᾳ δ' Ἐλένᾳ σ' ἐΐσ[κ]η ]κες ] . ις θνάταις· τόδε δ' ἴσθι̤ τᾷ σᾷ ] παίσαν κέ με τᾶν μερίμναν ]αις' ἀντιδ[ . . . . πό]θοις δὲ̤ ]    τί̤η̤ν̣ ]τας ὄχθοις ]αιν παν]νυχίσ[δ]η̣ν

Notes:

• In l. 1, there is a dot next to the top of the rho, possibly part of an epsilon; the lambda has left only two dots because a hole cut the rest off, and in fact that could well be a delta; the following gamma is next to the edge, so pi and eta are also options; hence the readings ΗΛΠ, ΗΔΗ by others; I followed Edmonds, hence my reading, with appropriate critical noation;
• In l. 3, the initial TI do look like part of an eta, given that the top of the tau crosses the iota at about half of its height; of course, an eta there is inmetrical, hence my reading choice; the delta and omega are partially superimposed, and ΔΟ is another option, but I disregarded it by metrical reasons;
• The two little dots opening l. 7 hardly suggest any reading; the theta and iota are both half-cut by a hole in the papyrus, which does look like it could allow one more letter; or perhaps the thing I am seeing as a trace of iota is just a quirk of how the ink of the tau was set down;
• The first letter of l. 8 is just a vertical stroke, presumably the right half of a pi; the mu has lost its left half and part of its right half, leaving only a caret-like thing midway through the line height;
• The first letter in l. 9 is a recognizable albeit half-cut alpha; a lambda perhaps? Nah… the final epsilon has left barely anything, oh wait, there is the middle stroke, a dot at the top, and part of the leftward belly;
• L. 10 consists of what is probably an annotation, perhaps a gloss; it does look strikingly similar to Edmonds' completion, which I adopted;
• L. 11 starts with a nu which has a barely visible slanted stroke and no right vertical stroke, then a clear upsilon with part of its vertical stroke lost, a clear chi and iota, the top of possibly a sigma, a lost letter, a pseudo-vertical stroke which actually slants and then straightens followed by another vertical stroke making an eta, and a clear nu.

P.Sapph. Obbink: a(n almost) new poem, and a revolution to a P.Oxy. 1231 poem

The image is theoretically copyrighted like that of the P.GC.s. However, I found it online. I guess it is copyrighted to Obbink, who gave his name to the papyrus? Whatever the case, I leave you with only the link, and do not embed the image. As shown in the image, these are actually two fragments, which split after l. 9 – that is, l. 9 is the last of the top fragment, and the second one begins with l. 10. Here comes the transcription.

ΑΛΛΑΪΘΡΥΛΗϹΘΑΧΑΡΑξΟΝ€ΛΘΗΝ ΝΑΪϹΥΜΠΛ€ΑΙ·ΤΑΜ€Ν̣ΟΙΟ̣Μ̣ΑΙΖ€ΥϹ ΟΙΔ€ϹΥΜΠΑΝΤ€ϹΤ€Θ€ΟΙϹ€ΔΟΥΧΡΗ ΤΑΥΤΑΝΟ€ΙϹΘΑΙ· ΑΛΛΑΚΑΙΠ€ΜΠΗΝ€Μ€ΚΑΙΚ€ΛΗ€ϹΘΑΙ ΠΟΛΛΑΛΙ̣ϹϹ€ϹΘΑΙΒΑϹΙ̣Λ̣ΗΑΝΗ̣ΡΑΝ €ξΙΚ€ϹΘΑΙΤΥΙΔ€ϹΑΑΝΑΓΟΝΤΑ ΝΑΑΧΑΡΑξΟΝ· ΚΑΜΜ€Π€ΥΡΗΝΑΡΤ . Μ€ΑϹ·ΤΑΔΑΛΛΑ ΠΑΝΤΑΔΑΙΜΟΝ€ϹϹ̣ΙΝ€ΠΙΤΡΟΠωΜ€Ν €ΥΔΙΑ̣ΙΤΑ̣Ρ̣€ΚΜ€ΓΑΛΑΝΑΗΤΑ̣Μ̣ ΑΙΨΑ̣Π̣ . ΛΟΝΤΑΙ· ΤωΝΚ€ΒΟΛΛΗΤΑΙΒΑϹΙΛ€ΥϹΟΛΥΜΠω ΔÁΙΜΟΝ€ΚΠΟΝωΝ€ΠΑΡΗωΓΟΝΗΔΗ Π€ΡΤΡΟΠΗΝ,ΚΗΝΟΙΜΑΚΑΡ€ϹΠ€ΛΟΝΤΑΙ ΚΑΙΠΟΛΥΟΛΒΟΙ· Κ̣ΑΜΜ€ϹΑΙΚ€ΤΑΝΚ€ΦΑ̣Λ̣Α̣ΝΑ€ΡΡ̣Η ΛΑΡΙΧΟϹ,ΚΑΙΔΗΠΟΤΑΝΗΡ❲€ΝΗΤΑΙ,                                   Ν ΚΑΙΜΑΛ€ΚΠΟΛΛΗΑΝΒΑΡΥΘΥ̣ΜΙΑΝ̣Κ€Ν ΑΙΨΑΛÝΘ€ΙΜ€Ν· ⸎ΠωϹ̣Κ€ΔΗΤΙϹΟΥΘΑΜ€ω̣ϹΑϹΑΙΤΟ ΚΥΠΡΙ,Δ€ϹΠΟ̣Ι̣Ν̣ΟΤΤΙΝ̣Α̣+.+ΗΦΙΛ̣[ ]Θ€ΛΟΙΜΑΛΚ̣ΤΑ+. .+Θ . . Χ̣ΑΛ̣[ ]ΟΝ€ΧΗϹΘΑ ]ΑΛΟΙϹ . Μ'ΑΛ€Μ . Τ̣ω̣ . ΔΑ¨ϹΔ[ ]ΡωΛΥΙ̣ϹΑΝΤ̣Ι̣[ ]ΑΠΑ . [ . . ]ΑΜΜΟ . . ΡΟ̣[ ]ΝΘ̣€Ρ . +.+Ι̣

Ἀλλ' ἄϊ θρύλησθα Χάραξον ἔλθην νᾶϊ σὺ‹ν› πλ‹ή›ᾳ· τὰ μὲν̣ οἴο̣μ̣αι Ζεῦς οἰ δὲ σὔμπαντές τε θέοι, σὲ δ' οὐ χρὴ ταῦτα νό‹η›σθαι· ἀλλὰ καὶ πέμπην ἔμε καὶ κέλεσθαι πόλλα λί̣σσεσθαι βασι̣λ̣ήαν Ἤ̣ραν ἐξίκεσθαι τυίδε σάαν ἄγοντα νᾶα Χάραξον· κἄμμ' ἐπεύρην ἀρτέμεας· τὰ δ' ἄλλα πάντα δαιμόνεσσ̣ιν ἐπιτρόπωμεν εὔδια̣ι ‹γ›ὰ̣ρ̣ ἐκ μεγάλαν ἀήτα̣ν̣ αἶψα̣ π̣έλονται· τῶν κε βόλληται βασίλευς Ὀλύμπω Δαίμον' ἐκ πόνων ἐπάρωγον ἤδη περτρόπην, κῆνοι μάκαρες πέλονται καὶ πολύολβοι· Κ̣ἄμμες, αἴ κε τὰν κεφά̣λ̣α̣ν ἀέρρ̣η Λάριχος, καὶ δή ποτ' ἄνηρ ‹γ›ένηται, καὶ μάλ' ἐκ πόλλαν βαρυθυ̣μίαν̣ κεν αἶψα λύθειμεν. ⸎Πῶς̣ κε δή τις οὐ θαμέω̣ς ἄσαιτο, Κύπρι, δἐσπο̣ι̣ν̣', ὄττιν̣α̣ +δ+ὴ φίλ̣[ησι κωὐ] θέλοι μάλ‹ισ›τα +πά+θαν χ̣άλ̣[ασσαι; ποῖ]ον ἔχησθα νῶν σ]άλοισί μ' ἀλεμάτ̣ω̣ς δαΐσδ[η ἰμέ]ρ‹ῳ› λύ{ι̣}σαντ̣ι̣ [ ]απα . [ . . ]αμμο . . ρο̣[ ]νθ̣ερ . +.+ι̣

Reading notes now.

• In l. 2, the top of the sigma and left half of the mu are faded out, but no other options are present;
• Same line, the second omicron of OIOMAI is blotted out and the mu has lost its middle, hence the doubtfulness, given that a middle-less mu could be a nu;
• In l. 5, the eta is crossed out and the epsilon is written over it;
• In l. 6, the I of ΛΙϹϹ€ϹΘΑΙ is exactly the hole, and the papyrus is crampled from ΒΑϹΙΛΗΑΝ on, which justifies the underdots;
• L. 7 is where the jumping lines begin; what I mean is that ll. 7-12 are cut into two, the second part being furhter up than the first; that is why the theta in l. 7 looks like an epsilon and something weird, but once you realize that, there is no doubt: it is a theta;
• L. 9 is split halfway through a pi which ends up looking like two taus; meter excludes that; that rho does look like a gamma, but if you look closely it does hook down, and given that rhos in this papyrus do not close a loop but only hook down, that will definitely be a rho; the epsilon is definitely not visible;
• In l. 11, that tau is definitely not a gamma: the horizontal top is just too far over on the left for a gamma, though a gamma is what should be there; the alphas are uncertain because they look a hell of a lot like lambdas; the rho is faded out and right after the split in the line; the last two letters are one terribly faded out and confused, the second looks like a mu, but a nu is required, and the shape is like a nu with slanted second vertical stroke and accidental falldown from the top of that;
• L. 12 is messed up by some sort of crampling, giving those uncertainties and making the epsilon just vanish;
• L. 14 strangely has a correction without crossing out; I say strangely because it seems that the standard in this papyrus is to cross out and then correct, as in all other corrections;
• L. 15 has the first of 2-3 "low dots" which I interpreted as commas to avoid confusion with unreadable vestiges;
• The mu in that line (μάκαρες) has lost its right slanted stroke, but since the rest is visible it leaves no doubts;
• The kappa in l. 17 has barely visible slanted strokes in a darker region of the papyrus, hence its underdot;
• The κεφάλαν in that line has messed-up ΑΛΑ (partly faded, with papyrus being "pixelated", as is true for the € and Φ too), and ἀέρρη has the last rho barely visible and the eta visible, though 2/3 faint;
• The eta in l. 19 was corrected to AN after being crossed out;
• In l. 19 again, the AR in ΒΑΡΥΘΥΜΙΑΝ is partly faded but not open to doubts, the upsilon I underdotted is almost gone, and the u is half-gone (the right half is gone, the left vertical stroke is clear);
• The little symbol starting l. 21 is a koronis, indicating a poem's end; it looks like a tau in the original;
• The sigma in that line has lost its bottom to fading, the kappa has lost its vertical stroke but is recognizable, the omega is barely visible and the epsilon before it is also quite faint;
• Despite ΚΥΠΡΙ being rather messed up in l. 22, it is certain; another "low dot"/comma follows; the sigma is clear, the pi is weird but I don't have other interpretations, the OIN is almost gone, the NA in OTTINA is almost gone, the following letter is totally gone (hence the pluses), the lambda is faint and in a darker region of the papyrus;
• The kappa in l. 23 does look like one, but could be (and P.Oxy. 1231 fr. 16 proves it to be) an IC where the sigma conects to the iota; again, the pluses denote blank spaces in the papyrus with an estimation of how many letters they consisted in; couldn't use <> because that indicates corrections in the reordered text, nor [] because that indicates a lacuna and there is no hole in that spot of the papyrus, nor {} because that indicates letters that are to be taken away (e.g. the {I} further below), and that was my next option, besides indicating such vanished letters with §; the lambda is half-lost to a lacuna, hence the underdot;
• The sigma in l. 24 has lost the middle of its top and the end of its bottom, making it look like an iota and a dot to the top-left of the following theta;
• No idea why Obbink sees a sigma at the start of l. 25; taking that mini-C which looks like a faint appendix to the alpha as a sigma seems precarious to me; the I of ϹΑΛΟΙϹΙ has left but a dot; ΑΛ€ΜΑΤωΝ has lost the second alpha and the nu to faint ink, and the letters in between are hard to see – in fact, the omega looks like an omicron followed by something unreadable; the ¨ is an Ï that lost the iota and retained only the diaeresis;
• The spurious iota in l. 26 has lost its top, hence the uncertainty; the TI ending ΛΥΙϹΑΝΤΙ are barely visible, and whatever followed is just gone, faded out, hasta la vista, no traces; big blank space left by that, btw;
• I would read the start of the following l. 27 as ΑΠΑϹ, but the sigma is close to a lacuna(?) hence the vestige status, and the pi is uncertain because it also suggests TI as an alternate reading, given the top that goes to the left but not the right of the two vertical strokes; the first vestige suggests an upsilon but I prefer to leave it unreadable; the other one is two dots at line bottom; rho certain, omicron barely visible;
• The theta may be an epsilon, and indeed Obbink reads it as ε̣; a trace is left before the lacuna(?), which therefore only removse one letter completely; after it, probably an iota; actually, it's more like νεερ+. .+ι, where the pluses could wrap three letters; in fact, one can even see a small dot just before the blanked-out part, so maybe ]νεερ . +. .+ι;
• Obbink reads an extra .ερης in l. 27; now, Grenfell-Hunt doesn't have it in 1231 fr. 16, and I cannot for the life of me see anything more in the papyrus, so I just leave it with nothing.

P.Oxy. 2289, fr. 5: traces of the first stanza of the first poem from above

This one has an image to it.

]ΛΑ̣ Ϲ€Μ . ΛΛÁΪ ΑΙϹΥ Δ€Ϲ[ ỊC̣Θ[

]λα̣ σεμ . ἀ]λλ' ἄϊ ν]ᾶϊ σὺ[ν δὲ σ[ ισθ[

With that big of an overlap, it definitely belongs with the Obbink papyrus above. The epsilon in l. 2 may carry an acute: a trace of it is visible above the mu. After the mu, a little dot, not enough to suggest an alpha, whereas the leftward belly in l. 1 is enough. The iota in l. 3 lost its right dot (that of the diaeresis). The bottom is burnt out, but looking closely and having Obbink's papyrus in mind the visible traces suggest the given letters. The [] in the raw transcription mark lines that begin after or end before the point where other lines do. This is partially kept in the reordered one, though additions skew that a little. The below combined text puts ] and [ around the lines coming solely from this fragment.

First poem from P.Sapph. Obbink: combined text

Combining Obbink with this yields, coloring this blue and Obbink no-color, and adding a pi to the first line because this poem is in an alphabetically arranged sequence where poems begin with omicron or pi, a sequence found in both P.GC. and P.Oxy. 1231:

Π[
[Lines missing, 1 or 5]
]λα̣[
]σεμ .[
Ἀλλ' ἄϊ θρύλησθα Χάραξον ἔλθην
νᾶϊ σὺ‹ν› πλ‹ή›ᾳ· τὰ μὲν̣ οἴο̣μ̣αι Ζεῦς
οἰ δὲ σὔμπαντές τε θέοι, σὲ δ' οὐ χρὴ
ταῦτα νό‹η›σθαι·
ἀλλὰ καὶ πέμπην ἔμε καὶ κέλεσθαι
πόλλα λί̣σσεσθαι βασι̣λ̣ήαν Ἤ̣ραν
ἐξίκεσθαι τυίδε σάαν ἄγοντα
νᾶα Χάραξον·
κἄμμ' ἐπεύρην ἀρτέμεας· τὰ δ' ἄλλα
πάντα δαιμόνεσσ̣ιν ἐπιτρόπωμεν
εὔδια̣ι ‹γ›ὰ̣ρ̣ ἐκ μεγάλαν ἀήτα̣ν̣
αἶψα̣ π̣έλονται·
τῶν κε βόλληται βασίλευς Ὀλύμπω
Δαίμον' ἐκ πόνων ἐπάρωγον ἤδη
περτρόπην, κῆνοι μάκαρες πέλονται
καὶ πολύολβοι·
Κ̣ἄμμες, αἴ κε τὰν κεφά̣λ̣α̣ν ἀέρρ̣η
Λάριχος, καὶ δή ποτ' ἄνηρ ‹γ›ένηται,
καὶ μάλ' ἐκ πόλλαν βαρυθυ̣μίαν̣ κεν
αἶψα λύθειμεν.

P.Oxy. 1231, fr. 16: more to the other poem in Obbink

This is a P.Oxy. 1231 and is not in this image, hence I have no image of it, but I do have the Grenfell-Hunt transcription, so here goes.

]ΘΑΜ€ω̣[ ]ΤΤΙΝΑ̣[ ]Λ̣ΙϹΤΑΠΑ[ ]. ]ĀΛ€ΜÁΤ[ ]·ΓΟΝωΜ[ ]Ο̣Μ'ΟΥΠΡ[ ]ΑΙ ]Ϲ€́·Θ€Λω[ ]ΤΟΠÁΘΗ[ ]Λ̣ΑΝ·€ΓωΔ€Μ'[ ]ΝÓΙΔΑ ]·[.]Ϲ̣ΤΟΙϹ[. . .] . [ ]€̣ΝΑΜ[ ]€̣[

]θαμέω̣[ ὄ]ττινα̣[ μά]λ̣ιστα πα[ ]. ] ἀλεμάτ[ων ]· γόν' ὠμ[ ]ο̣μ' οὐ πρ[ ]αι ]σέ· θέλω[ τοῦ]το πάθη[ν ]λ̣αν· ἔγω δ' ἔμ' [αὔτᾳ τοῦτο σύ]νοιδα ]· [.]σ̣τοισ[. . .] . [ ]ε̣ναμ[ ]ε̣[

I also have a few reading notes by Grenfell-Hunt:

• A superseded note on integrating ll. 2-4 with a quote that, with Obbink, we know does not fit there;
• In l. 9, «ε of θέλω was corrected from λ»; why not indicate this in the transcription then?! Was the lambda written in the same spot (i.e. overlayed) as the epsilon, or was it written above it as happens in corrections?
• Ll. 11-12 are integrated with a quote from Apollonius, which overlaps with almost everything we have of those lines (only the λαν is new);
• In l. 13, «the supposed stop may be the vestige of a letter, e.g. ε»;
• In l. 15, «what has been taken for the tip of an ε is possibly a circumflex accent»; l. 15 is the last line, so instead of ]ε[ it may be reading ]^-[.

Second poem from P.Sapph. Obbink: combined text

I now combine this with Obbink. On the left, I indicate Obbink uncolored, and this P.Oxy. 1231 blue, while on the right I swap colors. The quote is in both cases indicated as a supplement.

Πῶς̣ κε δή τις οὐ θαμέω̣ς ἄσαιτο, Κύπρι, δἐσπο̣ι̣ν̣', ὄττινα̣ +δ+ὴ φίλ̣[ησι κωὐ] θέλοι μάλιστα πάθαν χ̣άλ̣[ασσαι; ποῖ]ον ἔχησθα νῶν σ]άλοισί μ' ἀλεμάτ̣ως δαΐσδ[ην ἰμέ]ρ‹ῳ› λύ{ι̣}σαντ̣ι̣ γόν' ὠμ[ ]απα . [ . . ]αμμ' οὐ προ̣[ ]ν ἔ̣ερθαι ]σέ· θέλω[ τοῦ]το πάθη[ν ]λ̣αν· ἔγω δ' ἔμ' [αὔτᾳ τοῦτο σύ]νοιδα ]· [.]σ̣τοισ[. . .] . [ ]ε̣ναμ[ ]ε̣[

Πῶς̣ κε δή τις οὐ θαμέω̣ς ἄσαιτο, Κύπρι, δἐσπο̣ι̣ν̣', ὄττινα̣ +δ+ὴ φίλ̣[ησι κωὐ] θέλοι μάλιστα πάθαν χ̣άλ̣[ασσαι; ποῖ]ον ἔχησθα νῶν σ]άλοισί μ' ἀλεμάτω̣ς δαΐσδ[ην ἰμέ]ρ‹ῳ› λύ{ι̣}σαντ̣ι̣ γόν' ὠμ[ ]απα . [ . . ]αμμ' οὐ προ̣[ ]ν ἔ̣ερθαι ]σέ· θέλω[ τοῦ]το πάθη[ν ]λ̣αν· ἔγω δ' ἔμ' [αὔτᾳ τοῦτο σύ]νοιδα ]· [.]σ̣τοισ[. . .] . [ ]ε̣ναμ[ ]ε̣[

P.Oxy. 1787, frr. 1-2&2a: line ends for a (or two or three?) poems

Here is an image of the big one, P.Oxy. 1787 fr. 1, and one of the small ones, fr. 2, both apparently removed from online:

I unfortunately have no image of fr. 2a, but with some luck, a scan of the fragment will be provided here. Fr. 1 links to a quotation to which fr. 2 also links, hence they are together. Fr. 2a… According to GH, «this fragment has been included on account of its similarity on both sides to the upper part of Fr. 1; but that it belongs here is not certain». Hopefully I will get my hands on a scan and see said "similarities". Here are my transcriptions:

]ΙΔ̣Α̘[ ]ḤΑ[ ] ̣́ΓΟICẠ[ ] . [.] . [    ~12 letter    ] . ΔĀ́X[]θΗΝ      5 ]C̣Ỵθ[.] . []ω̣[    ~9 letters    ]ÝTẠN ] . Κ̣θΟ[..]ĂΤΙ . [......]Ο̣ΙϹΑ ]Μ€́ΝᾹΤΑ . [] ̣́ΝΥΜÓΝϹ̤€̤ ]N̤ΙθĤΤΑΙϹṬ[.]MA̤[.]ΠΡ̣ΟΚΟΨΙΝ̣ ]ΠωΝΚÁΛΑΔω . AΠΑΙΔ€Ϲ      10 ]ΦΙΛ'ÁΟΙΔΟΝΛΙΓΥΡΑΝΧ€ΛÝΝΝᾹΝ[ ] . ΤΑΧΡΟΑΓĤΡΑϹH́ΔΗ ]ΝΤΟΤΡÍΧ€Ϲ€ΚΜ€ΛΑΙΝΑΝ ]Α̣ΙΓΟΝ . . [.]ỴΦ€́ΡΟΙϹΙ ]ΗϹθ'ΙCẠΝΕΒΡ . . ΙϹΙΝ      15 ] . ΛΑΤÍΚ€NΠΟ€ÍΗΝ ]ΟΥΔΥΝΑΤΟΝΓ€Ν€ϹθΑΙ ]Β̣ΡΟΔÓΠᾹΧΥΝÁYωΝ ] . ΑΤΑΓÂϹΦ€ΡΟΙϹΑ ] . ΝÝΜωϹ€́ΜΑΡΨ€[      20 ]ÁΤᾹΝÁΚΟΙΤΙΝ[ ] . Μ€́ΝΑΝΝΟΜΙϹΔ€Ι ]ΑΙϹΟΠΑϹΔΟΙ ]ΤΟΥ̂ΤΟΚÁΙΜΟΙ ] . ΟΝ . . [.] . ΓΧ€      25 ] . [...] . C .́ .[ fr. 2 ]T̤ỌΛẠ[ ] ̣ΠῙΝ[ ]ΦÍΛ€Ι ̣[ ]Κ̣̣ΑΙN̤[ fr. 2a ]Λ̣€ϹϹΑ[ ] ̣Π€́ΡỊ[ ] ̣€ΙΚ€ ̣[

]ιδ̣α̣[ ]η̣α[ φ]ύγοισα̣̣[ ] . [.] . [    ~12 letter    ].δ' ἄχ[]θην     5 ]σ̣υ̣θ[.] ̣[]ω̣[    ~9 letters     α]ὔτα̣ν ] . κ̣θο[..]ᾰτι . [......]ο̣ισα ]μένα τὰν [πολυ]ώνυμόν σ̤ε̤ ]ν̤ι θῆται στ̣[ύ]μα̤[τι] πρ̣όκοψιν̣ ]πων κάλα δῶρα παῖδες      10 ] φίλ' ἄοιδον λιγύραν χέλυνναν[ ] . τα χρόα γῆρας ἤδη ἐγένο]ντο τρίχες ἐκ μελαίναν ]α̣ι γόνα δ' [ο]ὐ̣ φέροισι ]ησθ' ἴσα̣ νεβρίοισιν      15 ἀ]λλά τί κεν ποείην; ] οὐ δύνατον γένεσθαι ] β̣ροδόπαχυν Αὔων ] . ατα γᾶς φέροισα ] . ν ὔμως ἔμαρψε[      20 ]άταν ἄκοιτιν [ ἐκφθ]ιμέναν νομίσδει ]αις ὀπάσδοι ] τοῦτο καί μοι κά]λον λέ[λ]ογχε      25 ] . [...] . σ .́ .[ fr. 2 ]τ̤ο̣λα̣[ ]ἐπιν[ ] φίλει . [ ]κα̣ιν̤[ fr. 2a κά]λε̣σσα[ ] . πέρι [ ] . εικε . [

Notes on the doubtful readings:

• In l. 1 of fr. 1, the delta is marked uncertain because it looks almost like an alpha, though I tend to agree with the delta proposal of GH; the uncertainty of the iota doesn't convince me: what else could it be? An upsilon having lost its top? Then it would be twice as tall as alpha and delta: hmm…; also, 404 combining left tack above not found :);
• At l. 3, the eta is half-cut; the nu proposal of GH doesn't convince me because that middle stroke is, IMHO, far too horizontal for a nu; nevertheless, I print an underdot;
• At l. 4, it does seem, if you zoom in enough, that something is joining the I and the C into a M, but looking at it again, I don't think so; the alpha is only visible through its belly, the slanted stroke having been cut off; all I can see of that uncertain iota is a little dot, and the acute accent, so I'm leaving it completely unreadable;
• At l. 5, the chi is half-visible, half in the hole; the hole does seem to suggest another letter before the theta though; the final nu is also half-cut-away; this is where the second protrusion of the scrap (first one being the one containing the letters hitherto copied) starts; this line of said protrusion seems to have been ignored by GH, and indeed there are only vestiges to be found there;
• At l. 6, we have a very doubtful bit; I see a potential round stroke (something like ̖́ ) which could actually be the right half of a chi with faded left and middle or perhaps a C, then a potential Y, and what looks like a theta (θ), with a weird dot below; I can see, however, how others may join the round stroke and the small stroke of the Y into an alpha (Ⲁ), then use the weird dot to complete the rest of the Y into a lambda (Λ), and finally… nope, no way that's a lambda, that's still definitely a theta :); then there is a one-letter hole, then some signs that hint at an unreadable letter, then a hole, then that w-like sign which is either the bottom of an omega or that of a nu connected to the previous letter; or it could be the bottom of an ΑΛΛ, alpha having only the slanted part, first lambda clearly visible, second lambda half-cut; for some reason, GH ignored these things; after the following hole, ÝT is clearly visible, then we have what could be the belly of an alpha or maybe half of some other letter (e.g. omicron or theta), then a hole, then what looks like half a nu; I do not think any other letter could fit between wannabe alpha and half-nu; that justifies the underdots and the difference from GH;
• At l. 7, again we have a bit that GH for some reason ignored; the kappa is only half visible and could be a chi; then the ATI, plus a little stroke which I deem too vestigial to even warrant an uncertain lambda, since it could be a nu; the next bit starts with either an epsilon having lost the bottom, or an omicron having lost the top-right closure; I leave my original reading ọ out of laziness, but I deem GH's ϵ̣ just as possible;
• At l. 8, the mu is pretty clear, despite having lost its left vertical stroke and half of its right one; the epsilon has partly faded so that it looks like an underdotted sigma (C); the nu is half cut away, which is why I only give an unreadable letter; not quite sure if the following bit starts with something with an acute accent or a macron; certainly it couldn't be the omega we would want for the completion πολυ]ώνυμον, because the stroke is backslash-shaped, whereas omegas end in a stroke that is more like a rounded slash (\ vs /); alpha is possible, but so little doesn't seem to warrant an uncertain reading; the end is very uncertain: I'm not sure if that's a single C split over the hole or two letters, which would give a half-cut sigma and possibly an epsilon; I give double-underdotted C€ because the acute accent above the preceeding omicron seems (IMO) to suggest an enclitic like σε;
• At l. 9, the starting N is double-underdotted because it's half-cut, but a reasonable guess; the first tau has lost half of the vertical stroke but is certain; the other tau is uncertain because it's right at the border and the wannabe-horizontal stroke is a bit slanted, enough to allow the hypothesis of a half-cut nu (to my eyes); the mu is partly lost but enough of it survives to leave no doubt; not so for the alpha, where only the belly survives; the hole could fit TI or CI, but is not very wide, hence the one-letter width I gave it (and GH agree there);
• At l. 10, the reason GH are certain about the rho is that there is a descender to the left of what should be the letter; the reason I give a vestige is that, well, only the descender survived the hole that devoured the rest of the letter; those omegas are weird though: so compressed…
• At l. 12, the starting vestige is just a slightly slanted vertical stroke, decidedly not enough to justify a Ṇ as GH give, IMO;
• At l. 13, the starting nu is half-cut, but clear enough;
• At l. 14, you can clearly see a backslash-shaped stroke joined with a vertical stroke; this is what gave the Ṇ reading of GH; however, probably in the light of the other sources I'll discuss below, I read ẠỊ, interpreting that backslash as the slanted stroke of an alpha that has lost its belly, and the remaining vertical stroke as an I; indeed, if you zoom in like crazy, you can see a little dot just on the border of the papyrus, and that could be a trace of the belly; not sure whether I should keep that little high dot as a · or just discard it as accidental; again, on the basis of the other source, I don't think that punctuation sign is appropriate here; in the hole, I see vestiges of two letters; the first one looks like the belly of an alpha, but the other one looks like two dots, so is it a delta or is it the end of the slanted stroke of the alpha and the bottom-left corner of a delta? The upsilon could be taken for a tau, so underdot;
• At l. 5, the ϹΑ is certain, though I'm not sure how they split; either the sigma is just the bit around the hole, the protrusion of the papyrus into the hole bears the belly of the alpha, and the slanted stroke is slightly further right, or the sigma extends into said protrusion, and then the alpha has lost the horizontal stroke (aka half the belly); so the alpha will be underdotted; the vestiges are letters cut off almost entirely by the hole; the first one is part of a slash-shaped stroke, perhaps an iota, whereas the other is somewhat round, probably an omicron, and has another mark above it, possibly part of an acute accent on the iota; so probably this is ÍO;
• At l. 16, the starting vestige is a dot just above the bottom-left of the lambda; decidedly too little to warrant even a doubly-underdotted anything, let alone GH's single-underdot lambda;
• At l. 17, the omicron at the start is clearly recognizable despite being half-cut;
• At l. 18, the beta at the start is only half-extant (the top half is there, the bottom half is gone), hence the underdot;
• At l. 19, the first vestige (given as certain kappa (!) by GH) is just the upper portion of a slanted stroke, which could belong to a kappa, but also to a chi, and since the other source reveals ἔσχατα, well, a chi it is, but I can't say that here, so… vestige; oh yeah, there is also a wannabe connection of the vestige with the alpha, not sure if it's the lower part of a \-stroke from that chi or just a connection;
• At l. 20, the leading vestige is just a dot, no way I can give it as a certain omicron! The nu is not entirely extant but pretty clear; the upsilon carries a smooth breathing along with the acute accent, but I don't think I have a combining spirit to place above that upsilon, and trying to use the Greek keyboard for that would result in diacritics to the left of the upsilon, not over it; the final epsilon is almost cut off by the end of the papyrus;
• At l. 22, the vestige is just a dot I hadn't even noticed when transcribing;
• At l. 25, the first vestige is probably the right half of a lambda, but I cannot give it as an uncertain reading because it could also be an alpha having lost its belly; then we have the tops of two letters, the first one pointy so perhaps a lambda (or an alpha of the A-shaped kind), the second one rounder and with an acute accent, so probably an epsilon, not an omicron because it doesn't go back down; then we could possibly fit another letter in the hole, before reaching what is the far right of a letter, perhaps an omicron;
• My first transcription had actually ignored l. 26; now that I see it, I can barely see ink in the beginning vestige, then I see two vestiges, the second one accented and somewhat round, maybe ό or ϵ́, the first one is in fact a sigma, pretty certain, with a dot before it, and after these things another dot.
• L. 1 of fr. 2 is terrible; the leading tau is half-faded, the next letter is either an omicron or a squashed epsilon, the lambda is the only certain thing, and the alpha only has the belly; in fact, I originally read that as « ̣€̣Νθ»;
• The vestige in l. 2 is just a short slash, perhaps the top of an epsilon;
• The vestige in l. 3 is just a dot; I originally misread the line as « ̣ΓΛ€Ι»;
• The left of the kappa and the right of the nu are cut off, and there is a dot at the top of the line further right which could be the end of the nu's right upward stroke, or possibly another letter entirely; my original reading of the nu was «T ̣».

Let us now make the collage of these papyri. Unchanged color will denote fr. 1, cyan will be Athenaeus, and remembering fr. 2 is at the end and fr. 2a at the top should remove the ambiguity generated by having both of them denoted by red. A couple minor changes will be made to the transcription in the first few lines to better match the text I originally translated, which is probably B.A.'s text. Said changes are compatible with the notes I made.

κά]λ̣εσσα [     ο]ἶ̣δα [
] ̣πέρι̣[
] ̣εικε ̣[    ]η̣α [
φ]ύγοισα̣[
] ̣[.] ̣[    ~12 letters    ] ̣δ' ἄχθην̣      5
]χ̣υ̣θ[.] ̣[.]ω̣[    ~8 letters    α]ὔτα̣ν
]ἄχ̣θο[ς .]ατι ̣[......]ọισα
]μένα τὰν [πολυ]ώνυμόν σ̤ε̤
]ν̤ι θῆται στ̣[ύ]μα̤[τι] πρ̣όκοψιν̣
]πων κάλα δῶρα παῖδες      10
] φίλ' ἄοιδον λιγύραν χελύνναν[
] ̣τα χρόα γῆρας ἤδη
]ντο τρίχες ἐκ μελαίναν
]α̣ι γόνα δ' [ο]ὐ̣ φέροισι
]ησθ' ἴσα̣ νεβρίοισιν      15
ἀ]λλὰ τί κεν ποείην
] οὐ δύνατον γένεσθαι
] β̣ροδόπαχυν Αὔων
] ̣ατα γᾶς φέροισα
] ̣ν ὔμως ἔμαρψε[      20
]άταν ἄκοιτιν[
ἐκφθ]ίμέναν νομίσδει
]αις ὀπάσδοι
ἔγω δὲ φίλημμ' ἀβροσύναν [] τοῦτο καί μοι
τὸ λάμπρον ἔρ‹ω›ς ἀέλιω καὶ τὸ κάλον λέλογχε      25
] ̣πιν[    ] . [...] . σ .́ .[
]φίλει ̣[
]κ̣αιν̤[

P.Köln invv. 21351 and 21376: more to the above poem, and a doubt on the poem number

These fragments are put together because they are from the same piece of cartonnage, IIRC, and of course because of the blatant overlap with the above single Oxyrhynchus fragment. Here is the image:


First of all, naming problems. Looking at the image, one sees 4 fragments: a minuscule one on the left, a minuscule one on the right, a medium-sized one on the bottom-left, and the big one in the middle. Zooming in closer, one sees the medium-sized one is in fact 2 fragments, which seem joined by an isthmus, but are actually not joined at all. Reading Obbink and Hammerstadt, one then figures out the big one is two fragments, and zooms in and finds the joining line (probably). Then there is a question about the numbers. Here is what I came up with:

And here come the transcriptions:

fr. 1b ] ̣Ν̣Ο€̣[ ]Π̣ΗΜ̣[ ]ΦΑΤ̤[ ]Λ̣[ ]ΙΥ[ fr. 1 ] . ΟΥ[ ]€̣ΥΧΟ̣Μ̣[ ] . ΝΥΝΘ̣Α̣Λ̣[.]Α̣[ ] . €ΡΘ€Δ€ . ΑϹΠ€ . []ΟΙ ]Ο{Π}ΝΕΧ . Ι̣ϹΑΙ . ΕΡΑϹωϹ̣[.] Ο̣ΙΚ€̣Ν      5 ] . ΟΙ€̣Ν̣ΑϹΝΥΝ€ΠΙΑϹΕΟΙϹΑΝ ]ΛΥΓ . ρ̤Α̤Ν[..]ΚΕΝΕΛΟΙϹΑΠΑΚΤΙΝ[ ]Λ̣Υ̣Ν̣ . ΑΝ . ΑΛΑΜΟΙϹΑ€̣ΙΔω ]Κ . Λ̤Π̣ωΝ̣ΚΑΛΑΔωρΑΠΑΙΔ[.] . ] . ΦΙΛ̣ΑΟΙΔΟ̤Ν̤ . Ι̤Γ̣[.]ρΑΝ . [      10 ] . ΟΤ̣[.]Ο̣ΝΤΑ[...]ΑΓΗρΑ[ ] . . ΝΟΝΤΟ . . . ΧΕϹΕΜ[ fr. 1a ] . ΔΗ[ ]€ΛΑΙΝ . . [ fr. 2 ]ΒΑΡΥϹΔΕΜΟ ̣ΥΜ[.]ϹΠΕΠΟΗΤΑΙΓ[...]ΔΟΥ ̣[ ]ΤΑΔΗΠΟΤΑΛΑΙΨ ̣P̣EΟΝΟΡΧΗϹθΙϹΑΝ ̣ΒΡΙΟΙϹΙ. ] ̣̄ΤΑϹΤΕΝΑΧΙΖωθΑΜΕωϹΑΛΛΑΤΙΚΕΜΠO̤Ε ̣[ ] ̣ ̣Η̣ΡΑΟΝΑΝθΡωΠΟΝΕΟΝΤΟΥΔ̤ΥΝΑΤΟΝΓΕΝ̤Ε̤[ ]ΑΡΠ[.] . [.] . ΤΙθωΝΟΝΕΦΑΝΤΟΒΡΟΔΟΠΑ ̣Υ̣[      5 ]ΒΑΜΕΝΕΙϹΕϹΧΑΤ ̣ΓΑ ̣ ̣[ ]ΕΟΝΑΛΛΑΥΤ̣ΟΝΥΜωϹΕΜ[ ]ϹΕ[....]Αθ[.]ΝΑΤΑΝ[ ]ΜΥ̣θωΝ ̣[.]ΥΤ̈Ο̣ΥΕ ̣[ ]` ̣ ̣γε´[ ̣ ̣ ̣]ακ̣[ ̣ ̣] ̣[      10 fr. 2a ]Τ̣Ο̣[ ]ΝΤ . [ ]Υ[] . [ ] . ΤΤ̤Ο[ ]Α̣ΤΡ€̣Ι̣[ ]Ọ́ΘΡΙ . [ ] . ΝΑ[ inv. 21376 ΚΑΙΓ[..] . . [...] . . . . . [ ]ΕΡωΙΔ̣Ε̤[..]ΑθΕΙϹΑΝ[ ΕΟΝΤ ̣[..]Α̤Λ̤Ο̤ΝΚΑΙΝ[ ΧΡΟΝω̣Π̣[..] ̣ ̣ON̤Γ̤Η̤ΡΑ[ ΨΙθΥΡΟΠΛΟΚΕΔΟΛΙ̣Ε̣[ ἐπίβουλεπαῖ⟦β̣οτ̣ο ̣⟧[ ἑταῖρεἀφέρπω̣:δ[ [ ̣] ̣ ̣ ̣ ̣ ̣ ̣ ̣ ̣ ̣ ̣:(?) ̣ ̣[ [ ̣] ̣ ̣ν̣:(?)ἄπν̣ουςπρ ̣[ [φ]ά̣ο̣ς      15 ἀστέρωντε[ [τ]ὸ̣πυριφεγγὲςἀελ̣[ [ ̣] ̣πᾶςἀκούω:θ ̣ ̣[ Οἰά-] [γ]ρ̣ουκόρονὈρφέακ̣[      20 [ἑρ]πετὰπάντακ[

fr. 1b ] . ν̣όε̣[ισαι ] πῆμ[α ἔ]φατ̤[(ο) ]λ̣[ ]ιυ[ fr. 1 ] . οὐ[ ] ε̣ὐχο̣μ̣[ ] . νῦν θαλ[ί]α̣ ] . ερθε δὲ γᾶς περ[ίσχ]οι ]ον ἔχοι̣σαι γέρας ὠς̣ [ἔ]ο̣ικε̣ν      5 ]ζοιε̣ν̣ ἆς νῦν ἐπὶ γᾶς ἔοισαν ] λιγύρ̤α̤ν [αἴ] κεν ἔλοισα πᾶκτιν[ χέ]λ̣υ̣ν̣ναν̣ θαλάμοισ' ἀε̣ίδω ἰο]κόλ̤π̣ων̣ κάλα δῶρα παῖδ[ε]ς ] . φιλ̣άοιδο̤ν̤ λι̤γ̣[ύ]ραν . [      10 ] ποτ̣' [ἔ]ο̣ντα [χρό]α γῆρα[ ἐγ]ένοντο τρίχες ἐ‹κ› μ[ fr. 1a ] ἤδη[ μ]ελαίναν [ fr. 2 ] βάρυς δέ μ' ὀ θῦμ[ο]ς πεπόηται γ[όνα] δ' οὐ . [ ] τὰ δή ποτα λαίψηρ' ἔον ὄρχησθ' ἴσα νέβριοισιν ] ̣̄τα στεναχί‹σδ›ω θαμέως ἀλλὰ τί κε‹ν› πο̤εί[ην; ] ἀγήραον ἄνθρωπον ἔοντ' οὐ δ̤ύνατον γέν̤ε̤[σθαι γ]άρ π[ο]τ[]α Τίθωνον ἔφαντο βροδόπαχυ̣[ν      5 ]βάμεν εἰς ἔσχατα γᾶς . [ ν]έον ἀλλ' αὖτ̣ον ὔμως ἔμ[αρψεν ]σε[....] ἀθ[α]νάταν [ ] μύ̣θων α[]ὐτο̣υ̣ργ[έ      10 ]` ̣ ̣γε´[ ̣ ̣ ̣]ακ̣[ ̣ ̣] ̣[ fr. 2a ]τ̣ο̣[ ]ντ . [ ]υ[] . [ ] . ττ̤ο[ ] Ἀ̣τρε̣ΐ[δ ]ό̣θρι . [ ] . να[ inv. 21376 καὶ γ[..] . . [...] . . . . . [ ] Ἔρῳ δ̣ε̤[..]αθεισαν [ ἔοντα [κ]ά̤λ̤ο̤ν καὶ ν[έον χρόνῳ̣ π̣[ό]λιον̤ γ̤ῆ̤ρα[ς ψιθυρόπλοκε δόλι̣ε̣ [ Stopped transcribing here, trusting Hammerstadt for the rest. After a morning spent transcribing the rest of the papyri from this section and the above, I was a bit fed up of destroying my eyes over the damn faint cut-off uncials, right?

Reading notes:

• In l. 1 of fr. 1b, the nu is half-visible and could be an eta, though the sequence ΗΟ€ seems rather unlikely, hence the choice of N, while the epsilon could be a theta;
• In l. 2, the pi is half-cut and could be a tau, tough the horizontal top doesn't seem to stretch to the right of the vertical stroke, thus rendering the tau option unlikely; the mu's right vertical stroce has faded, or maybe that letter is a nu; I tried to match it to a NOHM I saw somewhere, but I realize now that what I remembered was the νό transcription made by GH of the last line of fr. 1 above, where I have mostly nothing;
• In l. 3, I «originally misread the phi as half-eta, but that is definitely an oval and not a stroke, as the eta would require; tay [recte tau] is barely visible», says the annotation I left to my transcription;
• In l. 4, «perhaps there is a letter vestige after this lambda? perhaps it's an alpha and not a lambda?», says the annotation;
• In l. 2 of fr. 1, one can clearly see what could be either the top of a delta or part of a nu, but it does seem to connect to the following certain upsilon with a vertical stroke as a nu, and that closing bottom for a delta is more like a quirk in the papyrus now that I look back, though I originally did read a delta there; then one sees faint traces lower that make this look like an epsilon, and gives up and just follows Obbink 'coz WTF man :); after that, I see a clear chi which I by some magic read as an omicron originally; looking close, there are traces suggesting ομ, but very faint;
• In l. 3, the more immediate reading is ΝΥΙΘΑ ., the theta being very faint, and the last vestige suggesting a lambda or perhaps alpha; looking closely though, you see the iota has a slanted stroke going down-right just like a u, and then you see the right vertical stroke of the nu superimposed with the faint theta; at that point, the nu is doubtless, and the theta could still be an omicron because the roundedness is unquestionable but the middle stroke is gone; alphas and lambdas are hard to tell, so you underdot the alpha too; and of course, the older ΝΥΙ€Ν reading I gave is bogus;
• In l. 4, I had a big fantasy when I first transcribed this: «] . €Ρω̣ΤΙ̤ΔᾹ . ΑϹΤ€Φ[ -- the first vestige could be a nu, the omega-tau pair could be an omicron-tau, and the tau-delta leaves me stumped because it is bizarre but apparently I find no better option; oh wait: just saw what could be the faintest iota ever up there after the tau, maybe that's it; so faint, double underdot is due; the vestige after that could be a pi; it's hard to tell whether those two letters at the end are two (e.g. epsilon-phi) or one (eta), but I think it's two; perhaps a form of ἀστέφανος / ἀστεφάνωτος?»; needless to say, that was bogus; real note, first vestige could indeed be a tau, after the rho we have a theta with barely visible middle stroke that is easily taken for an omicron, then what could be taken as a tau with a faint-ass small iota below its right side but is actually much better explained as an epsilon where the curved stroke is in the hole, then the epsilon after the delta does indeed look like Ā because the curved stroke is faint but present, then the gamma is faint-ass but still not totally gone, though the vertical stroke may be a papyrus quirk hence the unreadable status, the letter at the margin is confusing because it looks like a weird alpha or a 4, honestly I'll just make it unreadable;
• «]ΟΠΝΕΥ . ϹΑΙ . ΕΡΑϹω . [.] . Τ̣€ . -- The omicron in l. 5 is barely visible», or rather, the omicron is barely visible, the nu is "blackboard bold" (!!) making it really look like ΠΝ so I'll delete it as an emendation, the omicron is wholly in the hole while the iota has left margin traces, no clue why a certain nu could be taken out of a vertical stroke with nothing else after it which is definitely an iota (hence my ἔχοισαι where Obbink has ἔχοισαν), the gamma has only left its top, the epsilon is not curved, the sigma is half-gone as are the following underdotted letters;
• In l. 6, « ̣ΟΙθΝ̣ΑϹΝΥḲΕΤΟΙΤΑϹΕΟΙϹΑΝ -- the uncertain nu is barely visible; the kappa could be a nu, since the middle is faded», or rather the first vestige is like an underscore which probably extended shortly beyond the margin and then bent up for a zeta, the epsilon is easy to take for a theta, the pi has a curved right vertical stroke so it looks like TC or TO, no idea how I got that original kappa though, the top of the gamma does extend a little to the left but far too shortly to be a tau;
• In l. 7, «ΛΥΓρ̤Α̤ΝρΙ[..]ΚΕΝΕΛΟΙϹΑΠΑΚΤΙΝ -- double underdots indicate almost invisible letters; the ρΙ before the hole could be a N, and in fact I'll go for this idea in the collage below; the Π can also be read as TC, as I originally did», or rather, the letter after the lambda is an iota and an upsilon superimposed, a gamma and a faint trace follow, then a barely visible rho, and alpha, a nu with a slight top-left loop leading to that ρΙ, which is definitely an incorrect reading because the rho's top would't connect with the iota's bottom, then a two-letter hole, and a lot of certainty;
• In l. 8, «Α̣Υω̣ΝM̤ΑΛΑΜΟΙϹΑΤΙΔω -- the first letter could be a delta; the omega is very faint; the first mu is even fainter», or rather a lambda with a trace in the middle making it possibly an alpha, the top of an ypsilon, half-faded away, a papyrus quirk or a nu, a blank, a possible alpha connecting to an upward stroke which could start a nu, and a blank; the possible alpha + start of nu combination is closed on top by what is probably a papyrus quirk, and hence looks like an omicron or theta;
• In l. 9, «the omicron in the wannabe ΚΟΛΠωΝ has vanished, leaving a blank space, whereas the lambda is barely visible and the rho [recte pi] is half visible and could even be an eta; the kappa of κάλα «is also hard to tell from folds in the papyrus»; the alpha of παῖδες is half-cut, and the rest of that word is left as mere vestiges of lettertops», yeah mostly right, though the kappa is clear and the preceding nu isn't, and the lettertop-word has a clear complete delta beside the pi;
• In l. 10, the vestige at the start «looks like half mu but we want a nu!»; the lambda in the wannabe φιλάοιδον is half-cut and could be an alpha, and the ON is left as mere lettertops; the vestige after that should be the lambda of λιγύραν, the iota being half-visible, the gamma being visible [though connected to the previous letter which makes it look like an upsilon] and the upsilon filling the hole; there is probably a vestige of the top of the chi of χελύνναν;
• In l. 11, the first vestige could be a pi; the first tau is half-cut but undoubted, and the same holds for the alpha of ONTA; the omicron in ONTA doesn't close, strangely enough; the last vestige could be a trace of an eta;
• In l. 12, the first "vestige" is actually a blank space which should contain a letter but seems to have no trace of ink left; the second one is a letter's bottom, like the three ones before the chi, bottoms of letters cut away by a hole, which should match the tau-rho-iota of τρίχες; the chi is in fact half-cut and could be a kappa; the kappa matching ἐκ and the alpha matching the last one of μελαίναν seem to be missing, since the last vestige of l. 12 is more like a nu than an alpha.
• L. 1 of fr. 1b has a half-eta (right half), and l. 2 has bottoms of the final letters which are unrecognizable and would in fact point to just a nu, and no alpha;
• In l. 1 of fr. 2, the first vestige is «a letter that has faded away completely»;
• In l. 2, the vestige is the bottom of a possible eta, I can see the right vertical stroke and that's it; the rho is half-cut; the epsilon is almost a sigma (especcially if you don't zoom like crazy); the omicron is barely visible; that final period is more of a vertical stroke, no idea what it should be;
• In l. 3, «the I in TI is almost like a fold in the papyrus; that KEM should definitely be KEN, but that is not a nu to my eyes; the omicron in POE is hardly visible if at all»;
• In l. 4, the eta is half-cut; ANθΡωΠON has a faded alpha; that delta of δύνατον is barely visible, if it's there at all; final NE so hard to spot;
• In l. 5, the second hole is supposed to be a wrong space (?); the preceding vestige is the top of a possible tau, and the next one is blots of ink which may possibly suggest an alpha but it's too precarious to be taken that way; that tau in TIθωNON is underlined or something but then it goes further down!; the chi in the wannabe βροδόπαχυν is hardly traceable, and the ypsilon is hard to tell from the dark papyrus;
• In l. 6, that hole (between this fragment and the one joined to it) is probably a wrong space, although neither da*atheisa or ax*theisa give me anything on Perseus; I thought of zapatheisa from diapasxo but that would involve a zeta and anyway doesn't exist, it would have to be zapathoisa but that is not an omicron, it doesn't close; the end of the line is on a dark bit of papyrus which makes it hard to see the vestigial letters, the last of which is cut away anyways; actually, δαμάθεισα would be the aorist passive participle of δαμάω, "overpower", but would involve a poetic licence in that the stressed vowel is long but meter requires it short;
• In l. 7, «the tau in AYTON is over a hole and I originally read that as a pi»;
• In l. 10, the omega is half-cut, the left vertical stroke of the nu is almost all cut away by the hole; then there are vestiges of a letter on both sids of the hole, then what looks like an upsilon having lost its left slanted top stroke, then what looks like an upsilon with diaeresis (YY? What sequence is that?), then what looks like Ϲ€ ̣, and this is the last line I transcribed;
• In l. 2 of fr. inv. 21376, that delta looks like a letter (delta, alpha, lambda) with an underdot in the papyrus; B.A. gives a phi, but that is not a phi to me, and the article agrees with me on this; the following A is barely visible and might be a chi; actually, considering the epsilons in this papyrus are't lunate but have the modern capital epsilon's shape, and looking at the papyrus better, I see ink traces that make an epsilon there more likely, hence I updated the transcription;
• In l. 3, «what we have of the KALON is almost all terribly hard to see»;
• In l. 4, there are traces of ink making the first vestige; the hole could contain two letters or only one with a little "kerning" around it; the double-underdotted letters (especially the first two) are faint and very doubtful;
• In l. 5, the theta looks almost like an omicron; the first lambda is half-cut, the kappa is partly faded; the iota is half-faded, the epsilon is… where? Basically cannot see it now…; and this is the last line I transcribed.

Combining P.Oxy. and P.Köln

It is now time to make a combied text. The left column integrates the previous Oxyrhynchus collage with the Köln material, using blue for fr. 1, pink for fr. 2, and purple for fr. inv. 21376, keeping the non-common lines from Oxyrhynchus. The right column, on top of exactly the same integration, keeps the lines from Köln, thus using cyan for fr. 1b, and leaves the end of the fragments alone because it is not by Sappho so we don't care about it.

κά]λ̣εσσα [     ο]ἶ̣δα [ ] ̣πέρι̣[ ] ̣εικε ̣[    ]η̣α [ φ]ύγοισα̣[ ] ̣[.] ̣[    ~12 letters    ] ̣δ' ἄχθην̣ ]χ̣υ̣θ[.] ̣[.]ω̣[    ~8 letters    α]ὔτα̣ν ]ἄχ̣θο[ς .]ατι ̣[......]ọισα ]μένα τὰν [πολυ]ώνυμόν σ̤ε̤ ]ν̤ι θῆται στ̣[ύ]μα̤[τι] πρ̣όκοψιν̣ [ὔμμες πεδὰ Μοίσαν ἰο]κόλ̣πων κάλα δῶρα παῖδες σπουδάσδετε καὶ τὰ]ν φίλάοιδον λιγύραν χελύνναν[ ἔμοι δ' ἄπαλον πρίν] ποτ̣' [ἔ]ο̣ντα χρόα γῆρας ἤδη ἐπέλλαβε λεῦκάι τ' ἐγ]ένοντο τρίχες ἐκ μελαίναν βάρυς δέ μ' ὀ θῦμ[ο]ς πεπόητα̣ι γόνα δ' οὐ̣ φέροισι τὰ δή ποτα λαίψηρ' ἔον ὄρχησθ' ἴσα̣ νεβρίοισιν [ταῦ]τα στεναχί‹σδ›ω θαμέως, ἀλλὰ τί κεν ποείην ἀγή̣ραον ἄθρωπον ἔοντ' οὐ δύνατον γένεσθαι καὶ γάρ π[ο]τα Τίθωνον ἔφαντο βροδόπαχυν Αὔων Ἔρῳ δ̣ε̤[..]αθεισαν βάμεν εἰς ἔσχατα γᾶς φέροισα ἔοντα [κ]ά̤λ̤ο̤ν καὶ έον ἀλλ' αὖτ̣ον ὔμως ἔμαρψε[ χρόνω̣ι π̣[ό]λιον̤ γ̤ῆ̣ρας ἔ[χοντ'] ἀθανάταν ἄκοιτιν[ ἐκφθ]ίμέναν νομίσδει ]αις ὀπάσδοι ἔγω δὲ φίλημμ' ἀβροσύναν [] τοῦτο καί μοι τὸ λάμπρον ἔρ‹ω›ς ἀέλιω καὶ τὸ κάλον λέλογχε ] ̣πιν[    ] . [...] . σ .́ .[ ]φίλει ̣[ ]κ̣αιν̤[

] . ν̣όε̣[ισαι ] πῆμ[α ]φατ̤[ ]λ̣[ ]ιυ[ [Several lines lost] ] . οὐ[ ] ε̣ὐχο̣μ̣[ ] . νῦν θαλ[ί]α̣ ] . ερθε δὲ γᾶς περ[ίσχ]οι ]ο ἔχοι̣σαι γέρας ὠς̣ [ἔ]ο̣ικε̣ν ]ζοιε̣ν̣ ἆς νῦν ἐπὶ γᾶς ἔοισαν ] λιγύρ̤α̤ν [αἴ] κεν ἔλοισα πᾶκτιν[ χέ]λ̣υ̣ν̣ναν̣ θαλάμοισ' ἀε̣ίδω [ὔμμες πεδὰ Μοίσαν ἰο]κόλ̣πων κάλα δῶρα παῖδες σπουδάσδετε καὶ τὰ]ν φίλάοιδον λιγύραν χελύνναν[ ἔμοι δ' ἄπαλον πρίν] ποτ̣' [ἔ]ο̣ντα χρόα γῆρας ἤδη ἐπέλλαβε λεῦκάι τ' ἐγ]ένοντο τρίχες ἐκ μελαίναν βάρυς δέ μ' ὀ θῦμ[ο]ς πεπόητα̣ι γόνα δ' οὐ̣ φέροισι τὰ δή ποτα λαίψηρ' ἔον ὄρχησθ' ἴσα̣ νεβρίοισιν [ταῦ]τα στεναχί‹σδ›ω θαμέως, ἀλλὰ τί κεν ποείην ἀγή̣ραον ἄθρωπον ἔοντ' οὐ δύνατον γένεσθαι καὶ γάρ π[ο]τα Τίθωνον ἔφαντο βροδόπαχυν Αὔων Ἔρῳ δ̣ε̤[..]αθεισαν βάμεν εἰς ἔσχατα γᾶς φέροισα ἔοντα [κ]ά̤λ̤ο̤ν καὶ έον ἀλλ' αὖτ̣ον ὔμως ἔμαρψε[ χρόνω̣ι π̣[ό]λιον̤ γ̤ῆ̣ρας ἔ[χοντ'] ἀθανάταν ἄκοιτιν[ [Poem not by Sappho omitted]

P.Berol. 9722, fol. 5 (LP p. 15): "O Atthis"

This time we have an awfully faint papyrus. First of all, the numbering is nonsense. I align to Voigt, who has fol. 1-5 and then fol. 5 verso ed. Zuntz. Those are, in Lobel-Page, respectively p. 12, p. 12, p. 14, p. 14, p. 15, ed. Zuntz. The actual right way would be to split this into three parts (a second fragment was joined in on the left, and a third one on the right after Edmonds), split the next poem also into three for the same reason, split the poem after that into two (you can clearly see a fragment splitting off on the bottom-right), and have 8 fragments, each with recto and verso, but apparently nobody does this. That said, here is the image:

And the transcription.

]ϹΑΡΔ̤ . [ ] . . Α̣Κ̣ΙΤ̣ΥΙΔ . [ ωϹΠ . [....] . ωΟΜ€ΝΒ̣ . [] . Α̣ω̣Ϲ̣ [ Ϲ€̤Θ€Α̣Ϝ̣ΙΚ̣€̣ . ΑΝΑΡΙ Γ . ωΤΑϹΑ̤Δ̤ . ΜΑΛΙϹΤ̣'€ΧΑ . ΡΕΜΟΛΠΑ      5 . ΥΝΔ€ΛΥΔΑΙϹ̣Ι̣Ν€̤ΝΠΡ̤€̤Π€ΤΑΙΓΥΝΑ̤Ι̣ . . . Ϲ̤Ι̣Ν̣ωϹΠΟ . Α€̣ΛΙω̣ ΔΥΝΤΟϹΑΒΡΟΔΟΔΑΚ̤ . Υ . . . ΜΗΝΑ̤ ΠΑΝ̣ΤΑΠ€Ρ€ΧΟΙϹΑϹ . ΡΑΦΑΟϹΔ€ΠΙ ϹΧ€ΙΘΑ̣ΛΑϹϹΑΝ . ΠΑΛΜΥΡΑΝ      10 ΙϹωϹΚΑΙΠΟ . ΥΑΝΘ€ . ΟΙϹΑΡΟΥΡ . Ι̤ . ΑΔ€ . ϹΑΚΑΛΑΚ€ΧΥΤΑΙΤ€ΘΑ Λ . ΙϹΙΔ . . ΡΟΔΑΚ̣ΑΠ . . . Ν ΘΡΥϹΚ̣ . ΚΑΙΜ€Λ . Λ̣ωΤΟ . . . Θ̣€̣Μ̣ωΔΗϹ ΠΟΛΛΑΔΕΖΑΦΟΤΤΑΙϹΑΓΑΝ . Ϲ̤ . Π .      15 ΜΝΑϹΘ . ΙϹΑΤΘΙΔΟϹΪΜ€̤Ρω Λ̣ . ΠΤΑ̣ΝΜΟΙΦΡΕΝΑ̣Κ[.]Ρ̣'ΑϹΑ . ΟΡΗΤΑ̣Ι ΚΗΘΥΙ̣Δ̣€̤Λ̤ΘΗΝΑΜΜΟ̤ . [.]ΒΟΟΙϹΑ̣ΗΤΟΤ̣ΟΥΟΔΟΥΔΟΟΥ ΝωΝΤ̤Α [ . . ] . ϹΤ . ΝΥ . [.] . ΠΟΛΥω̣Ϲ̤ ΓΑΡΥ€̣ΙΔ[ . . . ]ΑΛΟϹΠ[ . . . . . . ] ΝΤΟΜ€ϹϹΟΝ      20 ]Υ̤Μ̤Α̤ . [ . . . ] . Ν̣Ο̣Υ̣Κ̣Α . ΜΙΘ€ΑΙ̤Ϲ̤Ι̤ΜΟΡ ΦΑΝ€ΠΗ̤[ . . . ]ΟΝ€ΞΙϹ . ϹΘ̣ΑΙϹΥ[ . . ]Ρ̣ΟϹ€Χ . ϹΘΑ[ . . ] . ΝΙΔΗΟΝ ] . . . ΤΟ[ . . . . ] . ΡΑΤΙ ΜΑΛ[      ] . €ΡΟϹ      25 . ΑΙΔ[.]Μ̣[      ]ΟϹΑΦΡΟΔΙΤΑ ΚΑΜ̣ . [      ]Ν€ΚΤΑΡ€Χ€ΥΑΠΥ ΧΡΥϹΙ̣ΑϹ̣ . [      ]Λ̣Ο̣Ϊ̣Α̣ . . . . ] Π Α ΠΟΥΡ̣[      ] . Χ€ΡϹ̣Ι̣Π̣€Ι̣Θω ]Θ+ . . +Η̣Ϲ€ΝΗ      30 ] . ΚΙϹ ] . . . . . Ν . . ΑΙ ] . ϹΤΟΓ€ΡΑΙϹ̤ΤΙΟ̤Ν̤ ] . ΝΦΙΛΑΙ ]Υ̣Ϲ̣Τ . ΝΟΥΔ€ΝΟ[      35 ]€ΡΟΝΪΞΟ[

] Σαρδ̤ . [ πό]λλα̣κ̣ι τ̣υίδε [ ὤς πο[τ' ἐ]ζώομεν β̣εβά̣ω̣ς̣ [ σὲ̤ θέᾳ̣ ϝ̣ικ̣έ̣λαν ἀρι- γνώτᾳ, σᾷ̤ δ̤ὲ μάλιστ̣' ἔχαιρε μόλπᾳ      5 Νῦν δὲ Λύδαισ̣ι̣ν ἐ̤‹μ›πρ̤έ̤πεται γυνα̤ί̣- κεσσ̤ι̣ν̣ ὤς ποτ' ἀε̣λίω̣ δύντος ἀ βροδοδάκ̤τυλος ‹σελάννα› πάν̣τα πε‹ρ›ρέχοισ' ἄστρα φάος δ' ἐπί- σχει θά̣λασσαν ἐπ' ἀλμύραν      10 ἴσως καὶ πολυανθέμοις ἀρούραι̤ς ἀ δ' ‹ἐ›έρσα κάλα κέχυται, τεθά- λαισι δὲ βρόδα κ̣ἄπαλ' ἄν- θρυσκ̣α καὶ μελίλ̣ωτος ἀνθ̣ε̣μ̣ώδης πόλλα δὲ ζαφο‹ί›ταισ' ἀγάνας̤ ἐπι-      15 μνάσθεισ' Ἄτθιδος ἰμέ̤ρῳ λ̣έπτα̣ν ‹π›οι φρένα̣ κ[ῆ]ρ̣'̤ ἄσᾳ βόρητα̣ι κῆθ{υ}ι̣ δ̣' ἔ̤λ̤θην ἄμμ' ὄ̤ξ[υ] βό{οι}η τ‹ὰ› {ο} δ' οὐ {δοου} νῶν γ̤' ἄ[π]υστα νὺξ {[.] .} πολύω̣ς̤ γαρύε̣ι δ[ι'] ἄλος π[όρω]ν τὸ μέσσον      20 [ε]ὔ̤μ̤α̤ρ[ες μ]ὲν̣ ο̣ὐ̣κ̣ ἄμμι θέαι̤σ̣ι̤ μόρ- φαν ἐπή̤[ρατ]ον ἐξίσω- σθ̣' ἀι σὺ [ . . ]ρ̣ος ἔχησθ' Ἀ[δ]ωνίδηον ] . . . το[ . . . . ] . ρατι μαλ[      ] . ἔρος      25 καὶ δ[.]μ̣[      ]ος Ἀφροδίτα καμ̣ . [      ] νέκταρ ἔχευ' ἀπὺ χρυσί̣ας̣ . [      ]λ̣ο̣ΐ̣α̣ . . . ἀ]π' ἀπούρ[      ] . χέρσ̣ι̣ Π̣εί̣θω ]θ+ . . +η̣σενη      30 ] . κις ] . . . . . ν . . αι ] ἐς τὸ Γεραίσ̤τιο̤ν̤ ] . ν φίλαι ἄπ]υ̣σ̣τον οὐδενο[      35 ἴ]ερον ἴξο[μ

Reading notes.

• In l. 1, the sigma is clearly visible, and connected to the previous letter, of which nothing is left (unless one wants to argue for an upsilon where the right slanted stroke is what I see as a connection, but that's too doubtful), the alpha has left its belly and half its slanted stroke, so it is certain, the rho is one of the weird-shaped rhos of this papyrus, no doubt on that, the delta is barely visible, and then there is a trace of ink or perhaps a shaded region of the papyrus;
• In l. 2, there are traces of two letters, then a very faint letter that is either a lambda or an alpha (or a delta without its bottom horizontal stroke), the kappa and tau are barely visible, the final vestige almost looks like an epsilon but it is too doubtful even for a double underdot; oh and the upsilon is more like a V;
• In l. 3, the left side starts; the vestige is a rounded downward stroke suggesting an omicron or perhaps an epsilon; the vestige before the second omega may be a papyrus quirk; there is this strange white granulous substance highlighting the letters which might actually be creating them; I assume that is not the case; the beta is hard to trace: more than one position is possible; the following vestige may once again be a papyrus quirk as it's on the margin; after a hole which contained the rest of the epsilon, an unreadable vestige, an apparent ultra-stretched alpha's bottom, what looks like an omega-sigma sequence very near the margin with the sigma half-cut off, and a number of mutilated unreadable vestiges I didn't even bother marking, though the line may end with a nu;
• In l. 4, the epsilon is rather faint, the other epsilon looks like a crossed-out omicron, the lambda is gone almost entirely, the alpha has a faint belly closure and could be a lambda, and the first iota may be a sigma, though to be honest I have no strong arguments for that, given that the bottom curve is very slight and may be accidental, and the top is cut away; oh and the second iota has a horizontal stroke above it which could make me think it comes from a preceding digamma; the kappa has almost lost its upper slanted stroke, ending up looking like a lambda instead; actually that epsilon may be an alpha corrected into an epsilon by adding a curve above it; OK I decided for an uncertain digamma instead of an uncertain iota, because adscript iota is not used in this papyrus;
• In l. 5, the nu has vanished almost entirely, hence the vestige; the alpha is like an alpha and an epsilon superimposed, and the following delta is barely visible; then we have unreadable traces; the tau is half-cut; the iota is almost invisible;
• In l. 6, the nu is lost to a hole; the CI looks like a mu, but I venture the sigma was traced bottom-up and joined to the iota, creating a spurious slanted stroke; then there is an incredible mess which clears out at ΠΡ€ΠΕΤΑΙ, and the final iota is faint hence uncertain; that alpha in GYNAI is so faint… in the "mess", I barely see a pi, the lacuna is mu-shaped, and just before it, the right vertical stroke of what seems to be the nu may actually be a faint-ass epsilon, meaning CIN is harder to trace than I made it to be;
  
  yeah, EMPREPETAI is likely, but whoa the faintness…; in fact, looking at it again, the epsilon can kinda be seen, the nu (not a mu!) is basically certain, and the pi is certain;
• In l. 7, we have what could be the upper slanted stroke of a kappa, then a blank with possibly the faintest epsilon ever, then a trace, then a barely visible sigma, a faint iota, a nu whose right vertical stroke is hard to see or coincides with part of the omega, the epsilon of AELIO and the omega are pretty hard to see;
• In l. 8, the kappa is visible in two separate places, both times very faint; the tau has lost its vertical stroke, hence the vestigial nature I gave to it; the LOS is just too hard to trace to give any kind of certainty; the final alpha is partially visible, and barely at that; oh and that vestige on PO<T>, the ink is indistinguishable from papyrus quirks, hence it could literally be anything;
• In l. 9, the nu looks like a rho, though the rho doesn't make sense given that perrechois' should go with astra but cannot, or would be referred to the moon and weirdly placed between article and astra, and a nu could be the thing if we assume this was written in one smooth motion, starting the nu at the low point where the alpha ended; the tau is invisible due to a dark papyrus and possibly faint ink;
• The uncertain alpha is faint and could potentially be a lambda in l. 10; the vestige is a vanished epsilon;
• In l. 11, the first vestige does suggest a lambda, but it is too faint to write a lambda there; the second vestige is incomprehensible because it is faint and in the dark side of this papyrus; at the end, there are confused traces of ink which I cannot interpret; there is a trace over a hole, then a vertical stroke that's possibly an iota, and a little blot I cannot decipher;
• In l. 12, the vestige is close to being a blank space; the Τ€ΘΑ is faint and over a dark papyrus, but if we assume that what is visible isn't papyrus quirks, it's all certain;
• In l. 13, all vestiges are due to undecipherable faintness; faintness is also responsible for the uncertain kappa, whose slanted strokes are barely visible;
• In l. 14, the kappa is faint hence the underdot; the following alpha has vanished; the wannabe iota of μελίλωτος is barely visible and definitely not warranting a double underdot; the lambda is faint but visible enough; the C AN are gone; the theta is hard to see but there; the epsilon and mu are doubtfully traced in that I have like two ways of retracing them, one involving weirdness, and one involving a spurious ink trace between mu and omega;
• In l. 15, the omicron is pointy and delta-like, but in other deltas the slanted stroke extends a little below the bottom, which doesn't happen in this case, hence the only option left is an omicron; the following letter should be an iota, but the horizontal stroke above it is unquestionable, and extends on both sides, so it's definitely a tau; the sigma is partially barely visible; the pi is clear; the rest is garbled, though an alpha can kind of be suspected for the first vestige; at the end, there seem to be TWO iotas, as opposed to the ONE we want;
• At this point I've considerably lowered my certainty threshold for faintness, or the MNAC would be . . . . because faint ink and dark papyrus… try looking at it! We are in l. 16 btw; the epsilon in the first word seems to superimpose with the theta or have two possible tracings, so I played cautious; the one in the last word is barely visible but enough for a double underdot;
• In l. 17, the lacuna signals a place where literally nothing is visible; the preceding lambda is barely visible, and the following pi has a clear right half and an invisible left half that would interfere with what seems like a nu; the two uncertain alphas may be taken for lambdas; the rho is uncertain because it's barely visible; trying to retrace Λ€ΠΤΑΝ, I made the lambda single-underdot because it could be an alpha, the epsilon is just a trace, pi is clear, tau too, alpha could be a lambda, nu is clear, that is no pi but rather definitely a mu;
  
  
• In l. 18, I originally wrote «Ϲ_Α̤Τ̤Ο̣^Δ̣ΟΥΟΥ», and commented «the double-underdotted pair is barely visible; the final omicron is garbled and Lobel-Page gives it as "."; what follows is a mess that I will now try to explore; well wow; we have a blank space, then an apparent sigma, then what looks like a subscript (!) alpha, then a tau, then a superscript (!) delta, a square (!) omicron, an upsilon, and ANOTHER OY! double underdots: barely visible; the omicron could be an alpha; the delta looks more like a chi»; but this part wasn't nearly described as messily as it was; up to "the final omicron […] gives it as "."", that is still valid; for what follows, the situation is much, much worse; here is my retracing attempt, where I retraced all the signs I saw:
  
  
  And no, I am not joking; I so much wish I were though… anyways, both Edmonds' and Lobel-Page's readings seem possible; I choose to only read the upper parts in the reordered transcriptio, and to drop the extras at the end (as everyone seems to do anyway);
• In l. 19, I originally wrote «ΝωΝΤ̤ΑΠ . [ . . ]ΑΝΥΞΠΟ . [.]+. . +ΠΟ€Ϲ··» and commented «apart from the starts of the big holes that will go on to the end of the papyrus, we have the final letters which I will ignore because they don't fit the meter»; then I revised the transcription, realizing I had seen things at the start, and took into account the fact that, as you can kind of see in the above retracing, the omega of πολύως is a C continued like an omega, justifying LP's πόλυς there, but if taken as an omega, one looks for a sigma, which Edmonds didn't have (the part of the papyrus on the right below this line was another fragment found later, and is indeed sellotaped to this part), but which I can barely see there; also, there is a removed part I cannot explain, so perhaps I'm deforming the text, but this way at least I get a complete line; trying to fill that gap doesn't make the line fit into the stanza, AFAIK; the only possibility I recall is ἄπυστον ὔμοι πόλυς; oh by the way, the double-underdot tau near the start of the line is so faint I cannot tell if it's a tau or a gamma, but I choose the latter because a conjunction doesn't fit here, and probably because it's what I adopted back in the days;
• The end of l. 20 is very garbled; taking what is suggested as certain since no other option presents itself;
• My gosh you REALLY have to pull your eyes out and stick them to the screen to see those double-underdot letters starting l. 21! the single-underdot letters I originally read MAPEC; the other double-underdot letters seem like papyrus quirks;
• In l. 23, the theta is squished hence uncertain; the middle portion is a terrible mess, time to explore it (or try);
• There seems to be an actual blank line in the papyrus between ll. 24 and 25, maybe the beginning of a new poem, except it's in mid-stanza, so mystery…;
• L. 26 features a number of partially mutilated but undoubted letters: the O in OC, the ΦΡ; the vestige in the first portion suggests an epsilon;
• In l. 28, to get Campbell's word χρυσίας, I assumed the iota was closing the right side of the sigma, that what I had seen as an iota after the alpha was actually the sigma, and what I had seen as a sigma was a vestige; my original reading was ΧΡΥϹΑΙϹ; as for the end, this reading was adopted by Lobel's Σαπφοῦς μέλη, whereas Lobel-Page and Campbell read ΝΑΝ, but I find the latter rather more unlikely, given it would require either the assumption that the second nu had the slanted stroke going the wrong direction (thus looking like a Cyrillic И instead of a N), or the assumption that, after a slanted stroke that is now gone, and a vertical stroke that ended up slanted, the nu in question came back down with a spurious vertical stroke, both of which are pretty far-fetched if you ask me;
• In l. 29, the spaces seem to be actually there in the papyrus, a thing never seen before; ϹΙΠ is uncertain because it's hard to locate, the other iota is mutilated by a hole;
• In l. 30, the uncertain eta is on a dark region of the papyrus (a black burnt-looking one) so it's hard to tell it from the rest of the papyrus; the +..+ is a two-letter blank space where letters have probably faded away; this would be the last line with a beginning portion, if said portion weren't half cut away and hence totally illegible;
• The vestige in l. 30 is a letter whose left half is cut away; reading it as an A, even uncertain, seemed too precarious to me; further ink traces ignored by LP seem to be present after the sigma; I will follow LP's suit;
• In l. 32, the vestiges suggest the inmetrical . ΛΑΙΚΝΑΙΑΙ;
• In l. 35, the upsilon has left only its right slanted stroke, the rest being beyond the margin, and the sigma is hard to see; the omicron is over a hole and hard to trace, unless one wants to see a mini-omicron that doesn't touch the hole, but it seems too small to me to actually be the thing.

As I said in the beginning, there are three fragments here. The top one contains everything up to l. 19, the leftmost part of l. 20, and the middle bit, excluding the sigma which is divided between the top and the bottom-right, oh and it also contains the top of the leftmost part of l. 21. The bottom-left contains the bottom of the left part of l. 21, and the left parts of all that follows. The bottom-right contains the rest of the lines from 20 on. This justifies Edmonds' reading παρενρεοίσας, which is mostly suppleted, when the bottom-right showing (or pointing to) πόρων τὸ μέσσον wasn't joined yet. Here are the joint lines traced:



I bid you goodbye with the monsters of this papyrus: an unidentified monstrous entity on the left, and E.T. on the right.

P.Berol. 9722, fol. 2 (LP p. 12): "I want to have died"

Here is the image:



To highlight a problem, here is how the lines run:



Terrible, right? With that realization, here are transcription and reading notes.

Τ€ΘΝΑΚΗΝΔΑ̣ΔΟΛω̣ϹΘ€̣Λω ΑΜ€ΨΙϹΔΟΜ€ΝΑΚΑΤ̤[.]ΛΙΠΠΑΝ€Ν ΠΟΛΛΑΚΑΙΤΟΔ'€€ΙΠ . [ . . ] . ωΙΜωϹΔ€ΙΝΑΠΕΠ[ . . . . ] . Μ€Ν ΨΑΠΦ . ΜΑΝ̣Ϲ̣'Α€Κ . Ι̣Ϲ̣Α̣Π . ΥΛΙΜΠΑΝω 5 ΤΑΝΔ'€ΓωΤΑΔ'ΑΜ€ΙΒ̣Ο̣Μ·ΑΝ ΧΑΙΡΟΙϹ'€ΡΧ€ . ΚΑ̣Μ€̣Θ€Ν Μ€ΜΝΑΙϹ̤Ο̈ΟΙΙϹΘ[ . . ]ΓΑΡωϹ€Π€Δ'ΗΠΟΜ€Ν Α̣ΙΔ€ΜΗ . ΛΛΑ . € . [.] . Θ€Λω Ο̣ΜΝΑΙϹΑΙ . . . . . [ . . ] . €ΑΙ 10 Ο̣€̣ . [ . ] . . . . [ . . . . . . . ]ΚΑΙΚΑΛ'€̣ΠΑϹΧΟΜ€Ν ΠΟ̤[ 10 letters ]ΟΙϹΙωΝ ΚΑΙ[ 11 letters ]Ο̣ΚΙωΝΤ̣'ΥΜΟΙ Κ̣Α̣P̣ . [ 9 letters ] . . Ρ€ΜΟΙΠΑΡ€ΘΗΚΑϹ̣ Κ . . Π̣ . [ 9 letters ]Θ̣ΥΜΙΔΑϹ 15 ΠΛ€Κ[ 8 letters ]ΠΑΛΑΙΔ€ΡΑΙ ΑΝΘ€ωΝ€̣ . [ 5 letters ]Π€ΠΟΗΜΜ€̣ΝΑΙϹ ΚΑΙΠΟΛΛω̣Λ̣[ 6 letters ] . Ϲ . . Ρω[ . . ] ΒΡ€ΝΘ€Ιω . [ 6 letters ] . . . . ΝΝ̣ . . €ΞΑΛ€ΙΨΑΟΚ . [ 4 letters ] 「. ΙΛΗΪω」 20 ΚΑΙϹΤΡωΜΝ̤[ 6 letters ]「ΠΙΜΟΛ」 . ΑΚΑΝ ΑΠΑΛΑΝΠΑ . [ 6 letters ] 「. . .」 ṆωΝ̤ . ΞΙΗϹ̣ΠΟΘΟ̣[ 7 letters ] . ΝΙΔ̣ωΝ ΚωΥΤ€Τ . Ϲ[ 7 letters ]Τ̣€ΤΙΤ̣Ι̣ ΪΡΟΝΟΥΔΥ[ 7 letters ] . Ν̤ 25 . Π . €Τ'ΟΠΠ̣[ 7 letters ] . €ϹΑΠ̣€ϹΚΟΜ€Ν ΟΥΚΑΛϹΟϹ . [ 8 letters ] . ΡΟϹ [ 20 letters ]ΨΟΦΟϹ [ 19 letters ] . . . ΟΙΔΙΑΙ

Τεθνάκην δ' ἀ̣δόλω̣ς θέ̣λω Ἄ με ψισδομένα κατ̤[ε]λί‹μ›πανεν πόλλα, καὶ τόδ' ἔειπέ [μο]ι ᾮμ' ὠς δεῖνα πεπ[όνθ]αμεν Ψάπφ' ἦ μάν̣ σ̣' ἀέκοι̣σ̣' ἀ̣π{.}υλιμπάνω 5 Τάνδ' ἔγω τάδ' ἀμειβ̣ό̣μαν Χαίροισ' ἔρχεο κἄ̣με̣θεν μέμναισ̤' οἶ{ι}σθ[α] γὰρ ὤς ‹σ›' ἐπεδήπομεν α̣ἰ δὲ μὴ ἀλλά σ' ἔγω θέλω ὄ̣μναισαι τὰ σὺ λ[ά]θεαι 10 ὄ̣‹σ›' ἄ[μ]μες φ[ίλα] καὶ κάλ' ἐ̣πάσχομεν πο̤[ 10 letters ]οις ἴω καὶ [ 9 letters κρ]ο̣κίων τ̣' ὔμοι κ̣ά̣ρ̣ᾳ [ 9 letters ] πὰρ ἔμοι παρεθήκα‹ο› καὶ π̣ό[λλαις ὐπα]θ̣ύμιδας 15 πλέκ[ταις ἀμφ' ἀ]πάλᾳ δέρᾳ ἀνθέων ἐ̣ . [ 5 letters ] πεποημμέ̣ναις καὶ πόλλῳ̣ λ̣[ιπάρ]ως μύρῳ βρενθείῳ . [ 6 letters ] . . . . νν̣ . . ἐξαλείψαο κα[ὶ βα]σιληΐῳ 20 καὶ στρώμν̤[αν ἐ]πὶ μολθάκαν ἀπάλαν πα . [ 6 letters ] . . . ν̣ων̤ ἐξίης̣ πόθο̣[ν 6 letters ] . νίδ̣ων κωὔτε τις [ 5 letters οὔ]τ̣ε τι {τ̣ι̣} ἶρον οὔδ' ὐ[ 7 letters ] . ν̤ 25 ἔπλετ' ὄππ̣[ποθεν ἄμ]μες ἀπ̣έσκομεν οὔκ ἄλσος . [ 8 letters ] . ρος [ 20 letters ] ψόφος [ 18 letters μ]ελαοίδιαι

Notes.

• First of all, there is no trace whatsoever of a lone psi from an otherwise lost line, as Edmonds would purport; however, there are end-of-stanza marks that tell us clearly that we have lost at least a line at the beginning, since l. 2 is followed by one and all other stanzas are made up of 3 lines, not 2; l. 2 is undoubtful (modulo the following note) despite being half-cut away and having only its bottom half left;
• In l. 1, the alpha is uncertain because it's barely visible, and so is the epsilon, whereas the omega is uncertain because it looks like two letters instead of one, though I cannot think of any option for the interpretation of this omega as two letters;
• In l. 2, the double-underdotted tau suffers from dark papyrus and is thus hard to see;
• In l. 3, the epsilons are faint, but zooming in reveals them in all their glory, and also reveals a tilde extending to both sides of the middle one of three vertical legs, which means it's not ΠΙ but ΙΠ that was written there; zooming in also reveals the apparent €̣ΝΗ at the end is in fact definitely Μ€Ν; the first omicron may look like an alpha, but alphas in this papyrus do not have the modern capital shape, so it is no alpha, nor a delta, since the horizontal stroke is not at the bottom but in the middle, so it must be an omicron;
• In l. 5, it seems we have a spurious letter between Π and Y; ICA are mutilated and hence uncertain; ΠΦ are also mutilated, but the leftovers leave no doubt; the NC is uncertain because it's hard to split them apart;
• In l. 6, BO is hard to see, and needs crazy zoom to even be suspected; note also how the tau has a clear vertical (or rather slanted) stroke and the endpoints of the horizontal stroke, while the middle of that stroke looks like a fold in the papyrus, and other folds make it easy to read X instead of T there;
• In l. 7, the omicron is essentially gone, hence the vestige status; the alpha is mutilated and may be a lambda; the epsilon is incredibly faint; there are traces further to the right but I cannot identify anything;
• In l. 8, the sigma is double-underdot because it is a messy trace; the Ö is apparently to be interpreted as deleted: as Lobel-Page puts it, «litt[era] o punctis appositis deleta» (the letter o being deleted by the dots added [to it]); the double iota is certainly a scribal error (or a misinterpretation of the ink traces by me);
• In l. 9, the first letter could be a lambda; the ΛΛΑ has a first lambda that is rendered blatant by zooming in; unless my vestige is to be taken as spurious, the epsilon as a sigma, and what I saw as a vertical stroke is actually traces of an epsilon, we have a lacuna that would need a letter but doesn't get any; the epsilon in Δ€ can only be an epsilon, despite being close to a blot of ink;
• The uncertainty of the omicron in l. 10 is due to its mutilation; the vestige trail are letters whose tops alone resisted being cut off; I do not agree that «θεαι legi nequit» (θεαι cannot be read) at the end;
• In l. 11, the first letter could be a theta; the second letter can only be a sigma if we suppose the pretty evident lower part is a spurious trace of ink, so I agree with Lobel-Page that «OC legi nequit» (OC cannot be read); the uncertain epsilon is blotty and hence uncertain;
• In l. 13, the uncertain tau is hard to tell from papyrus quirks; I first thought it might be a delta, but then I saw the vertical stroke and reread as tau, but I cannot say it is certain; no idea why LP reads ]κ̣ιων when I have a certain kappa and a preceding doubtful omicron;
• In l. 14, the beginning is faint and on dark papyrus, so it's hard to see; LP mentions others reading ΚΑΡΑ there; I cannot see a rho in there; the final sigma is weirdly shaped and may be an iota; LP says «παρέθηκας possis» (you could [read] παρέθηκας), no certainty, and then says someone corrected it; actually, looking at it again, the beginning could either be my original ḲẠỊ . . , or, given the shape of rhos in this, ḲẠP̣ ., where the rho is made up of the former Ị and another trace;
• In l. 15, the uncertain pi only has its bottom, the uncertain theta only its above-backslash part (I hope that is clear, it's not quite the right half, it's what is to the right of a backward-slanted line), and the vestiges are unreadable traces; how LP has ΚΑΙΠΟ̣ is beyond me;
• In l. 17, ἀνθέων is clear enough despite the faintness, the following epsilon is barely visible hence uncertain, then we have a vestige and a hole; the two mus are weirdly-shaped hence doubtful, what follows is a garbled mess where I cannot split the letters convincingly, then the alpha is almost an omicron, and the rest is clear; wait though: I missed the NAIC further right, and what comes before is definitely not what I thought, this is πεποημμέναις, with only the second epsilon mutilated hence doubtful;
• In l. 18, the two lambdas only have the right slanted stroke clearly visible, but there is definitely no middle stroke, so they are lambdas; the omega and lambda after them are hard to see; it is hard to tell whether . ωϹΜΥ[. .] is to be read on the right, though that looks more like a rho than a mu, which is why I chose what I chose;
• In l. 19, the Θ€Ιω is faint but doubtless; the hole has about 6 letters, then there are 4 letter traces, a clear nu, and a possible other nu, and then two vestiges, the second one possibly yet another nu; looking at it again, I can kinda see a very doubtful Κ€ΙΜ€Νω right below the hole between l. 18 and l. 20, which could also be Κ€ΙΜ€ΝΟΝ, but I haven't been able to extract any word out of it yet, except for κεῖμεον/κειμένω which I believe doesn't suit the passage, and should anyway be κημ- in Aeolic (cf. Edmonds' καὶ στρώμνας ἔπι κημένα in l. 21);
• The end of l. 20 is a goddamn mess; Ϊω is pretty doubtless, and working one's way back from that one can see all the letters, but half the eta and half the lambda are joined by spurious ink which makes what on first sight seems like a pretty blatant theta, except it's most definitely not; the beginning can suggest ΚΑ . [, but that would take part of the kappa as an alpha and see a complete kappa where there isn't one; also, it is NOT ΛΙΨ; if you look closely, there is a clear epsilon after the lambda, and the psi is preceded by a vertical stroke which is an iota; though there are spurious traces of ink which make the psi seem like it has a different shape, but we have seen psis that are crosses in this papyrus, so that is what we are looking for;
• In l. 21, the nu is barely visible and the wannabe theta has the weirdest 8-shape ever; the CT is almost like a Π, but the bends in the C rule that out;
• In l. 22, the first vestige is a vertical stroke; the others are a garbled mess with mutilations going on; the omega is clear, the following nu is hard to tell from papyrus quirks, hence the double underdot; the first nu is hard to see but easier than the other one, hence the single underdot;
• The uncertain omicron in l. 23 is barely visible; the vestige after the hole is possibly the slanted right stroke of an alpha; the sigma is barely visible; the delta has lost its horizontal bottom, hence it could be a lambda;
• In l. 24, I originally read the T of TE as an upsilon; the final TI is barely visible; the vestige is in a hole, but an iota is likely;
• I'm not even sure those extra vestiges ending l. 25 are actually there, so faint they are; actually, there is a trace before the possible nu which suggests a tau, though that is of course not a valid combination;
• In l. 26, the first uncertain pi is hard to see, the second one has lost its top horizontal stroke and could be II;
• Last line starts with the tops of 4 letters, the last of which is an omicron, and the other three are unknown; I believe €ΛΑΟ is possible, but other don't, according to LP's critical note.

As a final note, you can see joining lines that divide this into 3 parts. The top and left part contains most of this. The smaller middle part contains the parts in 「」 and the bottoms of the line before the first such part. The bigger bottom-right part has the rightmost parts of the lines from the first 「」 on. Here are the lines:

P.Berol. 9722, fol. 4 (LP p. 14): The other Gongyla poem

Compared to the previous P.Berol.s, this is much easier to read and quite a bit shorter. Here is the image:



Transcription and reading notes:

ΤΟΥ[ Η̣Ρ'Α[ ΔΗΡΑΤ . [ ΓΟΓ'ΓΥΛΑ . [ ΗΤΙϹΑΜ'€Θ€[ 5 ΠΑΙϹΙΜΑΛΙϹΤΑ . [ ΜΑΪϹ̣Γ'€̈ϹΗΛΘ'€ΠΟ̣[ €Ι̣Π̣ΟΜωΔ€ϹΠΟΤ'€[ [ . ] . ΜΑΓΑΡΜΑΚΑΙ「ΡΑΝ」[ [ . ]ΥΔ€ΝΑΔΟΜ'€Π「Α̣ΡΘ'ΑΓΑ .」 [ 10 ΚΑΤΘΑΝΗΝΔ'Ϊ「Μ€ΡΟϹΤΙϹ .」 [ . . ΤΙΝΟΙϹΔΡ「Ο . Ο̣€ΝΤΑ .」 [ . [ . ]ΟΙϹ̣ΙΔΗΝΑ「Χ̣€Ρ̣」[ [ . . ] . Δ'€€ΑΙΔ . [ [ . ]ΙΝΔ€̣Υ̣ΟΜ̣[ 15 . . ΜΗΤ . [

Του[ ἦ̣ρ' ἀ[ δήρατ . [ Γογγύλα . [ ἤ τι σᾶμ' ἔθε[λες 5 παῖσι μάλιστα . [ Ἔρ- μαις̣ γ' ἔσηλθ' ἔπο̣[ρος εἶ̣π̣ο‹ν› ὦ δέσποτ' ἐ[ [Ο]ὐ μὰ γὰρ μακά{ι}ραν [θέαν [Ο]ὖδεν ἄδομ' ἔπα̣ρθ' ἄγαν [ 10 κατθάνην δ' ἴμερός τις . [ λωτίνοις δροσό̣εντας [ὄ- χ[θ]οις̣ ἴδην Ἀχ̣ερ̣[ [κὰ]δ δ' ἐ‹ς› Αἴδα [ [ . ]ιν δε̣ύ̣ομ̣[ 15 . . μή τι [

Uncertainties.

• In l. 2, the eta is terribly hard to see, because the papyrus darkens like it was burnt;
• The vestige in l. 3 is the bottom of a vertical stroke, perhaps an iota, but perhaps any other letter with a vertical stroke (tau, rho, pi, others);
• The sigma in l. 6 is very faint and barely visible; the first epsilon is indeed bearing a trema (diaeresis);
• €ΠΤΟΜ comes more natural in l. 7;
• In l. 10, the final vestige suggests the left vertical stroke of a nu with a leftward hook rather than some weird sigma, so Campbell screw you :); the alpha is, however, uncertain, in that the belly is very faint and the slanted stroke could be an epsilon;
• The vestige in l. 11 suggests an epsilon to me more than the alpha Edmonds would have there;
• The uncertain omicron in l. 12 is hard to locate, I'm not sure what I saw is actually it;
• The sigma in l. 13 seems to actually close into an omicron; letters in that line are basically all incredibly faint; the sigma is impossible to locate; the NTA is very faint but clear; the mutilation of ΑΧ€Ρ motivates the underdots;
• L. 15 has a garbled mess in the €Υ, the epsilon being hard to see and the upsilon being a possible tau; the final mu is barely visible.

The picture clearly shows these are in fact two fragments, and in fact if you zoom you can see sellotape joining the smaller one on the right to the bigger one on the left. The right-side fragment has the parts in 「」, sometimes with a letter being half on one and half on the other, in which case it is counted in the one most of it is in.

P.Haun. 301: first part of the Kleis poem

Image:

Transcription and reading notes:

] . ΘΟ̈ϹΑΓΑΡΜ€Γ€ΝΝΑ̣[ ]Φ̣ΑϹ€ΠΑΛΙΚΙΑϹΜ€Γ . ]ΟϹΜ̣Ο̣ΝΑ̣ΙΤΙϹ€ΧΗΦΟΒΑΙ[ ]Π̣ΟΡΦ̣ΥΡωΙΚΑΤΕΛΙΞΑΜ€[ ]€̣ΜΜ€ΝΑΙΜΑΛΑΤΟΥΤΟΔ̣ . [ 5 ] . ΛΛΑΞΑ̣ΝΘΟΤ€ΡΑϹ€ΧΗ Τ̣ΑϹΚΟΜΑϹΔΑΙΔΟϹΠΡΟ[ ]Τ€ΦΑΝ̣ΟΙϹΙΝ€ΠΑΡΤΙΑ[ ]ΝΘ€ωΝ€ΡΙΘΑΛ€ωΝ ]Ι̣ΤΡΑΝΑΝΔΑΡΤΙωϹΚΛ[ 10 ] . ΙΚΙΛΑΝΑΠΥϹΑΡΔΙω[ Μ̣ΑΟ̈ΝΙΑϹΠΟΛ€ΙϹ

] . θος ἀ γάρ με γέννα̣[τ σ]φ̣ᾶς ἐπ' ἀλικίας μέγα κ]όσμ̣ο̣ν αἴ τις ἔχη φόβαι[ς π̣ορφ̣ύρῳ κατελιξαμέ[να ἔ̣μμεναι μάλα τοῦτο δ̣ή 5 ἀλλ' ἀ ξα̣νθοτέρα‹ι›ς ἔχη τ̣α‹ὶ›ς κόμα‹ι›ς δάϊδος προ[φανεστέραις σ]τεφάν̣οισιν ἐπαρτία[ις ἀν]θέων ἐριθαλέων μ]ι̣τράναν δ' ἀρτίως κλ[ 10 π]οικίλαν ἀπὺ Σαρδίω[ν ] Μ̣αονίας πόλεις

And here come the reading notes, as usual.

• In l. 1, the final alpha is half-gone; the omicron actually has an trema (diaeresis) above it;
• In l. 2, the initial phi only has the circle left, the vertical stroke being a possible papyrus fold; the alpha following it is faint but visible; the papyrus has a little protrusion at the end of the line, on which a dot is visible, which makes the final vestige together with another dot under the right end of the gamma's horizontal stroke; I believe the line is complete at that vestige, but perhaps it would be opportune to add a closing [ to the line;
• In l. 3, MO look like Τω; the alpha in AI is weird-shaped, it could be a lambda because the horizontal stroke is gone, or is high up and merges with the tip;
• In l. 4, the pi is half-cut away, and the phi may be a psi because the circle is half-gone;
• In l. 5, the delta is half-gone, and the top looks like a circumflex; the starting epsilon is hard to see on the border, but one can see the horizontal stroke and hints of all the rest;
• In l. 6, the uncertain alpha is half-gone; overlaying with the far right of the C in PAC is an apparent vertical stroke which, upon further inspection, reveals itself as an incredibly marked fold or depression in the papyrus;
• In l. 7, the initial tau is half cut away;
• In l. 8, the epsilon has a little dot far to the right in the place of the horizontal stroke, so it could be mistaken for a C, but that dot excludes that option; the nu has a hole in the middle and could be a mu;
• Between l. 9 and l. 10 we have dots tracing a trapezium; no idea what that is supposed to mean; a coronis should be further left, this thing is below €ω and above the first AN;
• In l. 10, the iota has just the bottom left;
• In l. 11, the lambda has lost a little bit at the top but is certain; the following alpha is half-cut away, hence the uncertainty; there are several blots above the line, responsible in particular for the state of mutliation of ΛΑ; before the iota, a round stroke, probably an omicron, or a papyrus quirk;
• The mu in l. 12 has a hole in its middle and could be, like, a nu;
• I honestly cannot tell if there is further ink after l. 12 or not; I believe not: what I see looks like a continuous dark line, more likely a papyrus quirk than actual letters.

P.Mediol. ed. Vogliano: second part of the above poem

Image:



Transcription and reading notes:

] . [ ΠΑΙ . ΑΠ€ΙΟΝ€ΧΗΝΠΟ[ ΑΙΚ€̣[ . ]ωΠΟΙΚΙΛΑϹΚ . . [ ΤΑΥΤΑϹΤΑϹΚΛ€ΑΝΑΚΤΙΔ . [ ΦΥΓΑϹΑ̣ΛΙϹΑΠΟΛΙϹ€Χ€Ι 5 ΜΝΑΜΑΤ . ΙΔ€ΓΑΡΑΙΝΑΔΙ€ΡΡΥ . [ ϹΟΙΔ€Γω,ΚΛ€ΙΠΟΙΚΙΛΑΝ ΟΥΚ€ΧωΠΟΘ€Ν€ϹϹ€ΤΑΙ ΜΙΤΡΑΝΑΛΛΑΤωΙΜΙΤΥΛΗΝΑωΙ

] . [ †παι . απειον† ἔχην πό[λιν αἴ κε̣[ν] ἦ ποικίλας κ . . [ Ταύτας τᾶς Κλεανακτίδα[ν φύγας †α̣λισα† πόλις ἔχει 5 μνάματ' οἴδε γὰρ αἶνα διέρρυε[ν Σοὶ δ' ἔγω Κλέϊ ποικίλαν οὐκ ἔχω πόθεν ἔσσεται μιτράν‹αν› ἀλλὰ τῷ Μ‹υ›τ‹ι›ληνάῳ

• L. 1 is only there because the papyrus has a non-straight border with a trace of ink above the X of €ΧΗΝ; in effect, there is a trace of ink above the A of ΠΑΙ too, but that could be an acute on the alpha;
• The more natural reading of the AI in l. 3 is OY, but then you note that the OY in l. 8 is very different in shape, with the omicron not linked to the upsilon, whereas the AK in ΚΛ€ΑΝΑΚΤΙΔΑ shows an alpha very much like the purported omicron which joins to the kappa's vertical stroke in the same way as this purported omicron joins to the vertical stroke of the purported upsilon, and that said vertical stroke is indeed vertical whereas the upsilon in l. 8 is an actual Y, so you conclude that this is AI and not OY; the following letter is very weirdly shaped and hard to recognize; then we have a hole and an H that lost the bottom of its left vertical stroke and, combined with half the pi, produces an apparent omega, excluded by observing that the end of said omega is in fact the beginning of the pi;
• I definitely see a sigma between TAYTA and TAC, no idea why it is not acknowledged by LP; maybe LP takes that as the tau? In effect, my tau is pretty thin and the vertical stroke is half cut away; oh I'm talking about l. 4 by the way;
• The alpha of l. 5 is mutilated but certain; the following sigma is hard to see in the holey mess; the following alpha is doubtful because the mutilation makes other options possible;
• The vestige in l. 6 is a rounded arc which may be an omicron, but «vix negaveris etiam in litt[eram] α posse quadrare» (you could barely deny that [the traces] could also fit with a letter alpha); the upsilon is mutilated but certain, then there is a trace that cannot be assigned to a letter with any kind of certainty;
• The following three lines, as LP puts it, «proprio loco omissos in fine columnæ lineolis marg[ine] sin[instro] appositis subiunxit manus prima» (omitted from their own place were placed at the end of the column by the first hand with added lines in the margin), that is the three small horizontal lines on the left of those three lines indicate they were added at the end and forgotten in their own place further above;

P.Berol. 5006 recto: parts of a poem addressed to Charaxus

This is a fragment containing middle portions of 11 lines of a poem attributed to Sappho – it's Aeolic, and Alcaeus is excluded for some reason, perhaps because some words are reminiscent of "Cypris and Nereids" aka "Divine Nereids", which is Sappho because it references events we know took place in Sappho's life (Charaxus's adventures and the dishonor he brought on himself). There is a trick to reading this papyrus, which makes many otherwise very doubtful letters certain, and it is to trace the strokes from the other side of the papyrus in the more messy places. Once you can eliminate what you are certain belongs on the other side, what is left is much clearer. I don't know if the papyrus is particularly transparent or some weird magic took place as the photo was taken, but that surely constituted a problem in reading it. Image incoming.

Trascription below.

]ΔωϹΗΝ ]ÝΤωΝΜ€́ΝT€Π[ ]ÁΛωΝΚἌϹΛωΝ . [ ]´ΛΟΙϹΛÝΠΗϹΤ€́Μ[ ] . ÓΝ€ΙΔΟϹ ]ΟΙΔH́ϹΑΙϹ.€ΠΙΤ . [ ] .́ ΑΝΑϹΑΙΟ·ΤΟΓΑΡΝ̤[ ] . ΟΝΟΥ . O̤̓ÝΤω . [ ]ΔΙÁΚΗΤΑΙ· ]ΜΗ̣Δ̣[

] δώσην κλ]ύτων μέντ' ἐπ[ κ]άλων κἄσλων . [ τοὶς φί]λοις λύπης τέ μ[ ] . ὄνειδος ] οἰδήσαις. ἐπιτ . [ καρδ]ίαν ἄσαιο· τὸ γὰρ ν̤[όημμα τὦ]μον οὐκ ο̤ὔτω . [ ] διάκηται· ] μὴ̣ δ̣[

A couple notes:

• As already observed above, it is in some places very hard to separate what is on this side and what is visible from the other side;
• The tau in l. 2 is actually superscripted, probably a later addition to the text (a correction);
• The accents are all actually in the papyrus, breathings included; I was surprised to see the latter, given that it's the first papyrus I see breathings on; actually, the second one, if the transcriptions are reported in order of making, given that there are breathings on P.Oxy. 1787 fr. 1;
• The vestige ending l. 3 is pretty badly mutilated, but still hints at a sigma;
• In l. 4, there is half an acute accent on the margin; said accent definitely continued beyond the margin, above the vowel it rested on;
• Some horiontal strokes are barely visible, but held certain nonetheless;
• The mu ending l. 4 is certain because the left vertical stroke is evident, and what is visible suggest the "V part" of the mu was traced as an arc of which part is evident and part is barely visible;
• In l. 5, the starting vestige is merely a vertical stroke, so I wonder why Campbell gives a certain mu when all I see is an iota; maybe that triangle to the left, which looks like 100% papyrus bullshit, is partially ink? Meter certainly does't force a mu, since δι' ὄνειδος is very well possible; I leave it as a vestige because, well, my completion starts with a mu there (ἔ]μ' ὄνειδος);
• L. 6 actually features a period; at the end, after the tau, a high stroke can be seen leading into the margin, whence the vestige;
• In l. 7, the starting vestige is a minimal trace, with a very clear acute accent above it; before TO, we have like a double high dot, something like ⋱ but without the lowest dot; the very vague trace on the figer-shaped protuberance of the line below, with the minimal trace just before the crevice between papyrus body and protuberance, hint at a nu, which is terribly mutilated and hence doubly underdotted;
• In l. 8, like in l. 5, only a vertical stroke is left, but the sentence strongly suggests the completion I gave, hence the vestigial nature; oh and a minimal trace of the right slanting stroke of said mu is barely visible on that vertical stroke's very near left;
• The vestige in l. 9 is totally gone, and is followed by what looks like an omicron, and above that apparent omicron is a very clear smooth breathing, while the Y carries a caron, as if this οὔτω got TWO smooth breathings, one for each component of its starting diphthong; at the end, a vertical stroke and a slanting stroke leading to the margin, the latter of which could equally likely have bent back up just after the margin to form a mu, or continued downward to form a nu;
• The eta in l. 10 has left only the tops of its vertical strokes, and the uncertain delta almost looks like the top of a lambda more than a delta.

P.Oxy. 424: more of that poem

This is, unfortunately, not available online, but I want to give you the Grenfell-Hunt transcription to show why it joins with the above. So here goes.

]ΑΙϹ€Π̣[ ] . Α̣Ι̣[ ]ΚΟΥ[ ]Α̣Ι ] . ΑΖ€·[ ]Χ̣ΙϹ·ϹΥΝΙΗΜ[ ] . ΗϹΚΑΚΟΤΑΤΟ[ ]Μ€Ν ]ΝΑΤ€ΡΑΙϹΜ€[ ]ΗΦΡ€ΝΑϹ·€Ý[ ]Α̣ΤΟΙϹΜΑΚΑ[ ]

]αισεπ̣[ ] . α̣ι̣[ ]κου[ ]α̣ι ] . αζε·[ ]χ̣ις· συνίημ[ ] . ης κακότατο[ς ]μεν ]ν ἀτέραις με[ ]η φρένας· εύ[ ]α̣ τοὶς μάκα[ρας ]

No notes are given by Grenfell and Hunt. The overlap with the above fragment is blatant and overwhelming, so no question can be made that they join. The only doubt could be what happens in l. 10 of the above (l. 5 here), where no overlap happens. Given that δόαζε is a thing, the above fragment has μὴ δ and this one . αζε, it seems reasonable to assume we should join those lines to read μὴ̣ δ̣όαζε, as Edmonds did and, of course, I adopted for my translations.
And with the image at here I completely agree with GH.

Combining the two fragments above

Of course, we now proceed to combine these two, coloring P.Berol. uncolored and P.Oxy. blue.

] δώσην
κλ]ύτων μέντ' ἐπ[
κ]άλων κἄσλων . [
τοὶς φί]λοις λύπης τέ μ[
] . ὄνειδος
] οἰδήσαις. ἐπιτ . [
καρδ]ίαν ἄσαιο· τὸ γὰρ ν̤[όημμα
τὦ]μον οὐκ οὔτω . [
] διάκηται·
] μὴ̣ δ̣όαζε·[
]χ̣ις· συνίημ[
] . ης κακότατο[ς
]μεν
]ν ἀτέραις με[
]η φρένας· εύ[
]α̣ τοὶς μάκα[ρας
]

P.Berol. 5006 verso: line endings of an otherwise non-extant poem

Like with the first above fragment, distinguishing what is on each side is a pain at times, and even more so on this side than on the other one. Image incoming.



Transcription below.

]Θ̣€ΘΥΜ̣ΟΝ ]ΜΙ ΠΑ . Π̣ΑΝ ] . ÝΝΑΜΑΙ· ] ]ΑϹΚ€ΝĤΜΟΙ ]Ϲ̣ΑΝΤΙΛΑḾΠΗΝ ] . ΟΝΠΡÓϹωΠΟΝ ] ] . ΓΧΡΟḮϹΘ€ΙϹ ] .́ [ . . ]Ρ̣ΟϹ̣

]θ̣ε θῦμ̣ον ]μι πάμπ̣αν ] δύναμαι· ] ] ἆς κεν ἦ μοι ]ς̣ ἀντιλάμπην κά]λον πρόσωπον ] ] . γχροΐσθεις ] .́ [ . . ]ρ̣ος̣̣

A couple notes:

• In l. 1, the theta is hard to see and doubtful, but the traces do hint at one;
• The mu in l. 1 has only left its vertical strokes, and the middle could even suggest a nu;
• There is an actual space between MI and the first pi in l. 2; the first alpha has left only its belly, but that is enough to unambiguously identify it; the mu is lost in the jumble of ink and holes; the following pi is two vertical strokes; oh and the first pi is weird and faint;
• The first vestige in l. 3 looks like ∠ flipped vertically (i.e. around a vertical axis), so a delta is suggested, but the bottom part is faint, so a lambda is also possible, and even an alpha; the high dot at the end is visible on the hole margin;
• In l. 6, the leading sigma has a half inverted breve at the top, and a micro-trace at the bottom, so perhaps other possibilities are available, but C certainly is the most natural choice; the acute accent has actually been misplaced over the mu;
• In l. 7, the vestige is a little "pointy dot" on the protuberance about which cfr. the other side;
• The gamma in l. 9 is very very faint, but I still think it is certain; there is a micro-line, a very faint one, right in the corner before the gamma;
• The rho in l. 10 has left IMHO too little to be certain: only the top is left; same for the sigma; the omicron is clearer because it is almost all there.

P.Berol. 9722 fol. 3 (LP p. 14): Edmonds 84 vs. Lobel-Page 93 vs. my own reading

This is the back of the other Gongyla poem above. As such, two fragments, but I won't separate transcriptions. I will use the 「」 notation to identify what comes from the smaller fragment on the left in the raw transcription, and to identify the parts present in the Lobel-Page text in the reordered transcription. The division line is blatant. Let's see the image.



I'm sure you can all see that we're in deep shit with this guy. But to make this impression worse, here is what this looks like in Edmonds (left) and in Lobel-Page (right):

    

Yeah, I know. It's crazy. And the worst part is that most of both can actually be traced and justified from the papyrus, or rather parchment. Wow!
So, given the incredibly faint nature of this guy, I decided it was opportune to transcribe it by hand, distinguishing 4 levels of relative faintness of the signs whose shape I imitated: black for clear, blue for clear enough, red for rather faint, faint red for almost invisible. And the result is:




And now we discuss what those garbled signs could actually have been: reading note time! The raw transcription tries to maintain the faintness levels as discussed below, with black mapped to black, blue to dim gray, red to gray, and faint red to dark gray.

• Between the top-right corner and that burn-like thing, the margin of the papyrus has a bump at one point. On that bump, it looks like there could be an omega. However, all other lines end much further left, so I'll assume that is a papyrus quirk, though I transcribed it with two underdots.
• In l. 1 (marked LP1 in the hand transcription), I choose to ignore the red (ultra-faint) parts of the first letter group, which looks like an infinity with the red, and like IC̣ without it. In the second letter group, we clearly see an epsilon, and then we seem to see a rho. One must observe, however, that the rhos of this parchment have no horiontal lines usually, and this one has two. So to get a rho, we must ignore two lines, If we ignore the shorter vertical one, we get a gamma. Looking at it again, the placement and color of those lines is wrong. Ah no, it's the epsilon that's doubled. The bottom of the one I drew is the top of another red epsilon. The rho should all be black outside of the top horizontal stroke, which is blue, and the top of the longer vertical one, which should be red or faint red. With that, we dismiss the horizontal strokes as papyrus quirks, and go for P̣. The other letter group is too far to the right, and is dismissed as a papyrus quirk. In the end, we get ΙϹ̣[..]€Ρ̣.
• L. 2 is all gray because, while the apparent rho is mostly blue, the short vertical stroke, the presence of which makes that into a rho and not a gamma, is red. Actually, scratch that: we keep the rho dim gray, but with an underdot.
• L. 3 is an example of how looking harder can overturn your transcription. The first transcription is my original one, where a lot of little signs were collapsed to the uncertain mu, the big mess was read as an omicron perhaps, the black thing was interpreted as a kappa, and the alpha would have fit in with Edmonds, whose reading I considered likely. To justify Lobel-Page, I looked harder, and the ex-left-half of the mu was turned to an alpha by an extra blue closure, the other half is a mess could be a lambda plus quirks, the mess is a very weird alpha, the ex-kappa got an extra red stroke which made it a likely mu, the ex-alpha looked like an omicron linked to an iota, and a sigma appeared. I guess Lobel-Page was really going for +...+μοισ, but the plus notation is my invention and there is no conventional alternative for that. My transcription will then be ]ΑΛ̤Α̣ΜΟΙϹ[.
• For some reason, l. 4 is skipped by all my sources. It is pretty straightforward to make out παρθένων there, with a messy epsilon that will be underdotted, and some weird letter connection. That matches a line in Lobel-Page and Edmonds, more or less… except it's not the one we're expecting to match this with, but the one below it! And since we'll see l. 5 is LP&E's l. 4, this line here is indeed absent in those sources. Good thing it was easy to read.
• In l. 5, we easily recognize the ΑΛΛΙ at the start and the ΗΧΟΝ at the end. What about the middle though? Looks like ΑΝ€_εΧ. Now, the subscript epsilon has the wrong shape, so I dismiss it as a papyrus quirk. The ΑΝ makes a word with the preceding iota. the rest looks like garbage, and was probably a messup by the scribe, which he later corrected with a marginal ΗΧΟΝ, and we saw that. So I'll keep it the way it looks, to dispose of it in the reordered transcription.
• L. 6 is the last line of Lobel-Page, and l. 5 for both Lobel-Page and Edmonds. For reasons of line length, I deem that from ωϹΟΜ on there are only papyrus quirks, so I leave that in the images. Just pointing out ωσομ literally reads as "awesome" :). Apart from that, let's break this down: we have (1) a vestige, (2) a clear pi, (3) the rectangle, (4) a clear enough double epsilon, (5) a scribble. (1)-(2) are clear, the lower part of (4) which makes the epsilon double is red so we dismiss it as a quirk and this becomes a certain epsilon, (5) may be ΜωΛ, or perhaps the epsilon continued upwards with a connection to a following omega, which then itself had a downward connection to a nu, which ended in a something, giving us €ωΝ., or again maybe the epsilon went up and down, thus connecting to a nu, which continued into an omega, which continued down and was supposed to slant back up for a nu but something happened to the parchment? I'll go for that idea, giving me ΝωΝ with all double underdots. We still have to clear up the rectangle though. With the revised transcription, we seem to see an alpha connected twice with the following letter, which is a rho, then €ΘΗΡΘ, and obviously the theta over rho-theta I forgot to put there again. I guess I'll ditch that in the reordering, taking what follows as an editorial correction.
• The start of l. 7 is definitely ΝωΡ. Then we see an omega fused with a rho, another omega, a delta continuing into an omega, yet another omega, a nu, a wonderfully clear though weirdly slanted ΥΠ€, a strange letter we will discuss later, a clear ON, what seems like a doubled theta, and a clear M. What is going on here? Firstly, assume the very first omega is not a rho with weird omega-like connections, but an actual omega with a blot giving a fake rho. In that case, we cannot escape a sequence of two omegas, which is almost impossible. So I prefer (perhaps with some Edmonds bias) to assume the weird omega-like connections, and have Ρ̣ω there. The delta is clear. The first omega after the delta is subscript-like, so I'll treat that as weird connections again. In the end, we get the word ωρρωδων, as in Edmonds, with some underdots and faintness markers. After the ΥΠ€, it seems like we have Ρ' with an extra small vertical stroke which is fainter. However, rhos here don't look like Ps, so the apostrophe and the vertical stroke should combine into a rho, leaving us with a belly and a vertical stroke to discuss away. I'll go for papyrus quirk, and underdot the rho. Theta-mu could be possible, but perhaps after ON we just have another ON, for some reason, which got papyrus quirks. Nah, I'll keep theta-mu with underdotted theta and ditch it in the reordered transcription.
• The following three lines are the Scary Movie of papyrologists :). L. 8 is unbelievable. We start with an eta, and that's fine. What follows looks like hieroglyphs, until we meet a clear alpha-beta sequence and an apparent omicron turned to an alpha by the slanted stroke above it. Then there is a one-letter space where I just gave up. What follows seems to be Λ€ with fancy balls to make it look cooler. Then another one-letter space where I gave up. Then two clear epsilons and some kind of a triple epsilon of sorts. Then a 4-letter space I just gave up on, and then ϹΘ᷍ΘΙ. So at the moment we have ]Η††ΑΒΑ+.+Λ€+.+€€€̣+....+ϹΘ᷍ΘΙ, and it's time to investigate that locus desperatus. A psi may seem to be hiding in the mess between the eta and the |-ρ, but I'll be prudent and grant it only the vestige status. The |-ρ looks like an eta with the right vertical stroke having left its top somewhere and a fancy ball having appeared at the end of the horizontal one. I'll say underdot. Then there is what seems like a rho with the fancy ball gimmick and the top having faded more than the rest. Underdot? Yeah, why not. So we end up with . Η̣Ρ̣.
• L. 9 starts with a possible vestige which is uncertain because it's right on the margin, then what is probably a weirdly faded phi which I'll average to underdotted and red, the following faint red rho is clear enough, then some kind of hieroglyph, then |-A᷍P, then what seems to be Ψ_α with extras to both letters, then a psi with extras to the left and a subscript xi to the right, then apparently Ỳ^Θ, then I gave up. Looking again, it seems like I missed a |\ before the |- on first pass, but |\|- is a nu connected to the following letter. The weird pseudo-hieroglyph suggests only one thing: an alpha, where the horizontal stroke doubled and both partially faded so they don't get to the slanted stroke, and then an extra slanted stroke created a fake delta on the bottom-right of the alpha. Or perhaps it's a subscript delta with extras we're looking at? Ạ be it. So the start has come to be Φ̣ΡΑ̣ΝΑ᷍Ρ. What follows could be as I analyzed it before, but the vertical stroke and the subscript are faint red, so we may dismiss them as papyrus quirk and read that as a red superscript underdotted omega. This also allows us to separate AP as an elided form of ἄρα. At this point it comes natural to assume a lot of quirks follow, and then ΞΥ, getting us the start of Edmonds' ὠξυβόων just before the giving up point.
• L. 10 starts with a clear pi, then a (probably accidental) two-letter surrender space, then ΑΡΘ€ is clear despite all the weird quirks under the epsilon, then it seems we have a slanted eta with a pi above, will underdot the eta for weird slanting, definitely not a nu though, then +.+, then a clear epsilon, then what seems like a nu with an extra middle stroke, either a double nu or a single nu with papyrus quirk, and since the slanted stroke is one I'll go for the latter, then the rest is clear.
• Last line, l. 11. Πω€ is clear, a mu seems to stick into the epsilon, then a pi, then +.+, then €Νω€ seems clear but I don't think omega-epsilon is a possible sequence, so I'll take the horizontal stroke of the epsilon as a papyrus quirk and read C̣, then a little anchor which was probably a tau, then TA is clear enough, the following seems to be a connection to a nu which, after going up for the last time, goes back down with a stroke that fuses with an epsilon, and we end with a nu and a surrender space.

With that, here is the transcription:

]ΙϹ̣[..]€Ρ̣ ]ΑΡ̣Υ[. .]ω ]ΑΛ̤Α̣ΜΟΙϹ[ ] . ΠΑΡΘ€̣ΝωΝ ]Α̣ΛΛΙΑΝ€ΧΗ̣ΧΟΝ ] . ΠΑ̣Ρ€ΘΗΡΘΘΡΘ€Ν̤ω̤ . ]ΝωΡΡ̣ωΔωΝ̣ΥΠ€ΡΟΝΘ̣Μ ]Η . Η̣Ρ̣ΑΒΑ+.+Λ€+.+€€€̣+....+ϹΘ᷍ΘΙ ]「.̣Φ̣Ρ」Α̣ΝΑ᷍Ρω̣ΞỲΘ+    + ]「Π+..+ΑΡ」Θ€Η̣+.+€Ν̣€Π̣€Β̣Α̣+...    +ω ]「Πω€Μ̣Π+..+€Ν」ωϹ̣Τ̣ΤΑΝ̣€̣Ν+    +

]「ισ̣」[..]「ερ̣」 ]αρ̣υ[. .]「ω」 θ]αλ̤ά̣「μοις」 ] . παρθέ̣νων ] 「ἀ̣λλ' ἴαν {εχ} ἦ̣」χον ] . 「παρ{εθηρθθ}έν̤ω̤ν」 ]ν ὠρρ̣ώδων̣ ὐπὲρ ὂν {θ̣μ} ]ηψ' Ἦ̣ρ̣α βά{+.+}λε{+.+εεε̣+....+}σθ‹α›ι ἠ]ΰφ̣ρα̣ν ἄρ' ὠ̣ξυ{θ}+βο-    + ] π{+..+}αρθέη̣ι +.+ ἔν̣‹ν›επ̣ε β̣α̣+...     + ]πω ἔμ̣‹μ›{+.+}εν ὤσ̣{τ̣}τ' ἀν̣ε̣ν+    +

Final remarks. Why the beginning of l. 2 is absent in both E and LP is a mystery. Maybe just too faint. Edmonds skipped l. 1 to avoid having to complete it, I assume. Why l. 4 is skipped by both is another mystery, because no mention of it is found anywhere. L. 3 is discussed above. The eta in l. 5 does kind of look like an omega, so that might explain LP's ἔχω. After the last LP line, the papyrus darkens at the extremes, and the text in the middle is terribly faint, so LP probably stopped because (a) dark papyrus means more likely to misread, (b) too faint to see anything. Even I have trouble finding my letters again, so I can't really blame LP for giving up on those lines :).

P.Oxy. 1787 fr. 45: a book 4 colophon?

Here is the image, apparently removed from online:

I guarantee you that, despite the papyrus fragment extending well beyond this image, the rest of it has no ink whatsoever. Transcription:

ΣΑΠ[ Μ̂ε̤[

Σαπ[φοῦς] με̤[λῶν δʹ

One thing only needs comments: the superscript letter in l. 2. The above transcription is based on this retracing:

This shows the epsilon in its full glory. Other possibilities exist though. For example, the following retracing makes this an ᾶ:

This one, a more far-fetched one, shows a nearly-lost alpha in the lacuna, with the epsilon being an editorial correction to it written over it:

Given that the transcribers of "The Oxyrhynchus Papyri" vol. 15, where the fragment is found, give με[, I'd guess they had retracing one in mind.

P.Vindob. 29777a: two twoside fragments

I bump these together because so does Lobel-Page in inc. 27. We have 3 fragments: fr. 1 recto, fr. 1 verso, and fr. 2. The rectos are on the left, the versos on the right. As you can see, fr. 2 verso is blank, except for what seems to be a trace of ink on the margin which is too precarious to be read, and would only yield a single letter anyway. Here are the images:

Transcriptions:

Fr. 1 recto ] . ΔΔ€Κ€ΤΑΙΜ€́ΛΑΙΝΑ̣[ ]ωΝ̤ΑΧ€́ωΝ̣€ΠÁΥϹΘΗ[ ] . . . ΪΔΑΙΤ€̤Λ̤€̣€̣Ο̣Ι̤[ Fr. 1 verso ]ωΙΚω ].ΟΝ ].ϹΟΤ¯ Fr. 2 ]Α̣ΙϹΟ€ΜΑ€ . ΟΑΤ€́ΡΥ[ ] . . ΝΟΗΜΑ ]Μ€Ν̣.Α̇ΦΙΚ€́ΤΑ[ ] . . Ο̣Υ̣Ϲ̣ . . [

Fr. 1 recto κ]αδδέκεται μέλαινα̣[ πόλλ]ων̤ ἀχέων̣ ἐπαύσθη[ Ἀ]̣τρεΐδαι τε̤λ̤έ̣σ̣θη̤[ν Fr. 1 verso ]ωικω ] . ον ] . σοτ¯ Fr. 2 ]α̣ισοεμαε . οατέρυ[ ] . . νόημα ]μὲν̣ ἀφίκετ' ἀ[ ] . . ο̣υ̣σ̣ . . [

Reading notes

Fr. 1 recto

1. First letter only has a bottom trace, last letter is almost complete, though a bit too high.
2. I'm going with the hypothesis that the vertical strokes of the first nu have almost entirely faded, and a papyrus quirk makes the slanted one look like a chi. I'm not sure where exactly the second nu is, which is why it's underdotted.
3. The first epsilon is very faint, and its horizontal stroke seems to clash with the lambda, which thus looks like a weird alpha, hence the dottings. The two epsilons are very faint, so €Ϲ is not out of the question, and the omicron may well be a theta, and that final iota is barely visible, so it could well be part of an eta, meaning τελέσθη[ν is not out of the question.

Fr. 1 verso

The letters LP lacks (i.e. the first of l. 1 and the tau of l. 3) are very faint.


Fr. 2

1. The alpha has its top cut off, so I leave it some doubt, the vestige suggests to me a ϲϜ combination which I don't think is possible.
2. The first vestige is a slightly right-down-slanted stroke which could be the top of a tau, the second one looks like part of a capital A, in particular I think it's too pointy to be an omicron, though a lambda is unacceptable unless the first vestige is a vowel, and even then ΛΝ is a strange pair. The eta is easily taken for a mu, and the MA at the end is like a scribble.
3. The nu is actually followed by a dot, and its slanted stroke is essentially gone. Another dot is found on top of the alpha (the first one).
4. I will not describe the vestiges except by saying they are lettertops. So are the OYC too, but those seem more readable: I mean, the top of an upsilon is basically unmistakable, the sigma is almost complete, the omicron is strongly suggested by the rounding, though a sigma is also technically possible.

P.Oxy. 1787 frr. 35 + 40 + P.Oxy. XXI add. p. 136

Image:

I will start by treating this as a single entity and transcribing it, then I will discuss numbers and splittings.

] . . || ] .́ΛΒΟΝ || ] ΑΚΟÝΗΝ || ] . ÝΤΑ . || ] .́ΓΑΝ || Ν [.] .[ ] || ΓΟ̣[ ] || . [ ] || . [ ] || Ọ́ . [ ] || ÀI[ ]Ν || €Ν[ ] || Ϲ[ ]⨉ || €́[

] . . || ] ὄλβον || ] ἀκούην || ] αὔταν || ] .́γαν || ν[.].[ ] || γο[ ] || .[ ] || .[ ] || ὄδ[ο- ] || αἰ[ ]ν || ἐν[ ] || σ[ ]⨉ || ἔ[

|| separates the two columns of this (joining of) fragments. Reading notes:

• In l. 1, it is uncertain whether we have vestiges of two or one letter, though two seems more likely;
• In l. 3, both the alpha and the nu have left only vestiges, respectively a quasi-horizontal line and what seems like a top corner, which is why I leave them as vestiges and not underdotted letters;
• In l. 6, the nu might look like a mu, but that apparent upward slanted stroke is a papyrus quirk, or the curve at the bottom of the right vertical stroke is unexplained;
• In l. 7, I'm not sure if that's an omicron or an alpha; I leave that as an omicron because of LP;
• In l. 8, we have a left semicircle, which could be an omicron, an omega, or a sigma; not an epsilon or theta, though, because those have horizontal strokes crossing their semicircles and such strokes are absent here;
• In l. 10, the vestige is a horizontal line on the bottom, which could be a delta, a zeta, or maybe a xi;
• In l. 11, I read αἰ because I know the breathing cannot go on the alpha since the grave accent on the alpha signals a diphthong, and obviously in Aeolic the rough breathing doesn't exist;
• In l. 13, I don't know why LP gives a vestige when we have a semicircle that is clearly detached from the margin;
• In l. 14, I have no idea what that X is, and it's not a chi because that letter can't end a word in Greek; the letter might be construed as a digamma, hence the uncertainty, but I must say it would be a pretty weird digamma if it were one.

That said, the numbers. LP informs us that his 85(a1), 85(a2), and 85(a3) are the combination of P.Oxy. 1787 fr. 35, fr. 40, and P.Oxy. XXI add. p. 136. One would assume he lists the fragments top to bottom, but then Voigt omits a2 and a3, and says a1 comes from fr. 40. So that is fr. 40, and ends at our l. 3. If you look at the image, the rest is not physically connected, which is why LP says the space between fr. 40 and the rest is "incertum". You can also see where the other two pieces separate, but they join with a snug fit, so I'll count them as united. Finally, P.Oxy. XV (which I luckily have) tells me that fr. 35 is the bottom one, so the middle one is the addendum.
In terms of dividing poems, I will count fr. 40 separate from the rest because we have no joint, but then I'll have col. i of the other two and col. ii up to l. 11 where a coronis is found as a single fragment (with "fortasse explicit" at the end of col. i), and the rest of col. ii as another fragment, because a coronis means poem end.

P.Oxy. 2293 frr. 1(a)+1(b): three fragments

Image:



Let me first transcribe each column of each fragment according to the numbers I wrote, then I will discuss this nonsense numbering.

fr. 1(a1) col. 1 ]€ ] ]ΙΝI ]Ν ]7 ]7 fr. 1(a1) col. ii Ν[ ΝΑ[ €Ν . [ ΦΟϹ[ ΔΕθΥ̣Γ[ Π€Ιθω[ ΤΑϹΑΛΛΗ[ ΓΟΡ€Υ€· ÝΜ[ ΘΙΚΟΝ€Τ.[ ΑΜΜΙΑΓΓ[ ΔΑΙΜ+.+ΝΑ̣ . [ ΪΝΑΗ . . Π€ . [ Λ€Γ'ΟΥ̣ΜΜ€[ Μ€Ν+.+.Ϊ€[ ΔΗ€Π . Ϲ̣ . [ θ€Λ€ . Ο̣[ ΟΝΔΤ€Κ[ . ΔΥΝΑΤ[ Χ€ΡΡ€Ϲ̣[ ΤΪΔΙΑΝ[ ΠΡΟϹΤΗ[ Τ[.]ϹΠΡΟ . [ . . €ΡΥΓ[ ]Ο̣ΦΗϹ . [ ].[ fr. 1(a2) ]Χ̣ . ] . ΝΤΙ ] . €Ϲ€Ι ] . ΟΝΤωΝ ]ΚΥθ€ΡΗΑϹΤΡÓ ]Π . Η€ΝΑΛΛΟΙϹ ]ΔΙΤΗϹ€ΙΡΗΚ€ΤΗ[ ]ΗϹ€Φώ̣Ν€ΙΗΜ[ ] . €ΑΥΤΗϹΠ̣Ρ[ ] . +.+ θ̣€ΛΟΙϹΑ[ ]ΑϹΙΝ ¨̇Χ[ ]ΤΙΝΑ[ ] . Ο̣ϹΥΝ[ fr. 1(b) col. i ] . ] . . ΦΙ ] . ΤΟΥ ] €̣ΙΜΑΙ' ]ΚΑΙΚΑ ] . ϹΚΑΙ ]Χ̣θ̣ΟΥΝ[ ] . ΑΝΥ[ ] . ΑΤΟ̣[ fr. 1(b) col. ii . [ ΡωΧΟΥ[ ϹΑϹΓ̣€̣ΡΑϹ: . [ ΚΑΙΓΥΡΙΝΝ[ ΤΑϹΤΟΙΑΥΤΑϹ̣ . [ ΓωΤΟΚΑΛΛΟϹ€ΠΙ€Τ[ Μ€ΖΟΝΤΙΓΑΡΗΝ€Μ[ €ΙΝΑΙΚΑΙΑΡΕΤΗϹ̣ΠΟ̣[ ΛΑΜΗΠΟΤ'€Λ€Γ€ΙΟΤΙΟ[ ΚΑΛΛΙ+.+€ΥΦΗΜ€ΙϹθΑ[ ΜΟΙΖ€ΦΥΡ̣ωΠΝ€ΥΜ . [ ϹΟΙΔΑΝ+. .+ . ΦΟΡΗΤΟ[ ]Ν̣ΟΝΔ€ΚΑ . [ ]€̣· ΠΑΪΤΑϹΜ[ ]ΔΗ̣ΝΓ€ΓΡ[ ]ΥΠΟΑΝΔ[ ] . ΙΟΥΚ€ΥΝ̣[ ]ω̣ΑΡΡ€ . [ ]Χ̣ΗΤΙϹ[

]ε ] ]ι ]ν ] ] ν[ να[ εν . [ φος[ δὲ θυ̣γ[ατέρα Πειθὼ [ τας ἀλλη[ γόρευε· ὐμ[ θικονετ . [ ἄμμι ἀγγ[ελίαν δαίμ+ω+ν ἀ̣θ[άνατος ἵνα η . . πε . [ λέγ' {ο}ὔ̣μμε [ μεν +.+. ιε[ δηεπ . σ . [ θέλετο̣ [ ον τε κ[ . δύνατ[ χέρρες̣ [ τ' ἰδίαν [ πρὸς τὴ[ν τ[.]ς προ . [ πτέρυγ[ες ]ο̣φησ . [ ].[ ]χ̣ . ] . ντι ] . ες ει ] . οντων ] Κυθερήας τρό- θρέ]πτη ἐν ἄλλοις Ἀφρο]δίτης εἴρηκε τὴ[ν ]ησε φω̣νείη μ[ ] . ἑαυτῆς π̣ρ[ ] . +.+ θ̣έλοισα[ ]ασιν χ[ ]τινα[ ] . ο̣συν[ ] . ] . . φι ] . του ] ε̣ιμαι' ] καὶ κα- ].ς καὶ ]χ̣θ̣ουν[ ] . ανυ[ ] . ατο̣[ . [ἀγε- ρώχου[ς τὰς ἄγαν ἐχού- σας γέρας· . [ Καὶ Γυρίνν[ τὰς τοιαύτας . [ἐ- γὼ τὸ κάλλος ἐπετ[ μέ‹σδ›ον τι γὰρ ἦν ἔμοι [ εἶναι καὶ ἀρετῆς̣ πο̣[        ἀλ- λὰ μήποτ' ἔλεγε{ι} ὅτι ὄ [ κάλλ‹ε›ι εὐφημεῖσθα[ι μοι Ζεφύρ̣ω πνεῦμα [ σοι δ' ἀν+εμ+οφόρητο [ ]ν̣ονδεκα.[ ]ε̣· παῖ τᾶς μ[        πρὸς Ἀνδρομέ]δη̣ν γέγρα[πται ] ὑπὸ Ἀνδ[ρομέδης ] . ι οὐκ εὐν̣[ ]ω̣αρρε . [ ] χ̣ῆτις [

Reading notes

fr. 1(a1) col. ii:

• In l. 5, I downgraded the Y to underdot because the bottom is all that's left, so it could be a T or even a Ψ.
• In l. 6 there should really be a +.+ at the end.
• In l. 8, the M is half-lost, but the slanted stroke clearly curves back up before the margin, so a N is excluded.
• In l. 9, I have no idea why that would be a certain rho with a macron over the margin, and not an epsilon, but either way it looks weird, so I'm leaving it a vestige.
• In l. 11, the alpha is almost entirely lost to a hole, and could IMO also be a lambda, and the theta is just gone, leaving only a microscopic ink dot right on the margin, so why LP would have it as anything more certain than a vestige is beyond me.
• In l. 13, the upsilon was upgraded from vestige because it's pretty visible to me, though it could technically be a chi as well (then again, chi-mu-mu is not a valid sequence in Ancient Greek AFAIK).
• In l. 15, I have no clue why LP would want a lost letter between epsilon and pi when the space there is basically the same as between eta and epsilon. The sigma has been downgraded to underdot because, being so close to the margin, it could easily be the top of an epsilon IMO.
• In l. 16, the tau was downgraded to vestige because all that is left is a ¯, so it could be a tau, a zeta, or a gamma. The epsilon has been turned to an omicron because we only see a semicircle, but an epsilon could easily have lost its horizontal stroke to give this.
• In l. 17, the delta was strikethrough'ed and the tau above is what should have been in its place.
• In l. 19, the sigma was downgraded to underdot because it looks more like a weird epsilon, but the top part could be a papyrus quirk or some weird mark, leaving us with a sigma which sounds better here, because -es is better than -ee, and even assuming this is a third person, there wouldn't be -ee but just -e, unless that starts a new word of course.
• In l .23, the first two letters were downgraded from uncertain pi-tau to vestiges because the pi has barely anything left and the tau has just the top, though the connection to the epsilon excludes the zeta option here.
• The vestige of l. 25 is more or less lined up with the sigma in l. 24.
• Note that, between l. 2 and l. 3, and between l. 22 and l. 23, there are weird markings I'm inclined to view as koronides, which means lots of text divisions, and way more than one fragment…

fr. 1(a2):

• In l. 3, the first vestige is further left than the vertical margin, but the papyrus extends just below the line, and that is where the vestige is found.
• In l. 5, the eta looks like it was EI and then the left leg of the eta crossed out the epsilon.
• In l. 6, the vestige was downgraded from an underdot tau because I can barely see the horizontal top and nothing more. Two vestiges that don't exist were dropped at the beginning.
• In l. 8, the omega is hard to see and thus rightly uncertain, but the acute is pretty clear, I wonder why LP didn't have it…
• In l. 9, the pi is uncertain because, having lost the right leg, it could be a tau, which was in fact my original reading.
• In l. 10, there is a trace in the far left of the outstretching, then a blank spot in said outstretching, then a top semicircle which could be a theta but also other letters, an epsilon which I see as pretty clear and thus upgraded from underdot to certain, and then the rest is doubtless.
• In l. 11, there are actually three dots over a blank, one slightly higher than the others and sitting between them; I'd say that's probably an interline annotation.
• In l. 13, the vestige is a top corner, as if from the top-right of a mu, and the omicron is half-gone and could be a theta, hence the uncertainty, just like in LP.

fr. 1(b) col. i:

• Squint very hard to find the dot that gives you l. 1.
• In l. 2, one vestige is a descender left of the margin that goes into l. 3, and the other one looks quite like an iota.
• In l. 3, I believe the above-mentioned descender hits another vestige, and then we have the TOY.
• In l. 4, the epsilon has lost its top and could be a sigma, hence the downgrade to underdot. The I' at the end could IMO be a rho, but I don't know of a word EIMAP (there is HMAP, but I couldn't justify this vowel change).
• In l. 6, the eta was downgraded to a vestige because a barely-visible right leg is all that's left, and the ".[ ]" was dropped because where in the world is that vestige, when a gap like that would lead far from the papyrus's edge?
• In l. 7, the chi is half-gone, hence the underdot and my original reading as sigma, the theta is also half-gone (the top is lost), I read it as an omicron originally, or maybe an epsilon, so underdot again.
• In l. 9, again we have an omicron that's half-gone, hence the underdot.

fr. 1(b) col. ii:

• In l. 3, gamma and epsilon were downgraded to underdot (actually only gamma) because the gamma could be a tau, since we don't see the point of intersection between vertical and horizontal part, and the epsilon has lost its top and could be a sigma.
• In l. 5, the last sigma is guessable even without vestiges, but uncertain because half-gone.
• In l. 6, another deletion happens with the iota with epsilon written above it.
• In l. 8, the sigma was downgraded to underdot because there is a mini-dot which could make it an epsilon, while the omicron stayed underdot because the top is lost.
• In l. 10, the eta is certain because if you look hard enough you can see a trace of the horizontal stroke which clears away every doubt (including that it might be an iota as I thought originally).
• In l. 11, the rho was downgraded to underdot because the vertical leg is gone and the rest could also be an omicron or theta, and the final alpha was downgraded to a vestige because it could be several other letters with what it has left.
• In l. 12, I see the nu, then a two-letter-wide blank gap, then a vestige, which was probably the Ọ of LP, though it has far too little to guess at its form IMO.
• In l. 13, the nu is half-gone, hence the underdot.
• In l. 14, the epsilon was read as certain sigma by LP, but there is a middle stroke which makes it more of an epsilon IMO.
• In l. 15, the eta was an omega to me originally, hence the downgrade to underdot, and the delta is upgraded to certain because, thinking about it, it couldn't really be another letter.
• In l. 17, I gave in to LP in starting with a vestige instead of an upsilon, but not in the certainty of the nu, which I downgraded to underdot because it could definitely also be a mu.
• In l. 18, the omega is uncertain because I can only see it by following the margin, which is not a very trustworthy method of deciphering traces.
• In l. 19, the chi is half-gone.

The supplements are taken from this Italian anthology, which I'll come back to while discussing the numbers.

Here is what I originally wrote about these numbers in the "Rest of Sappho" draft.

For starters, the numbering of the single fragments in P.Oxy. 2293 "fr. 1" makes no sense at all. I have a picture, and there are three fragments. One has microscopic endings of a column and more consistent portions of several lines of a second column. Let's dub it fr. 1(a). A second, smaller fragment contains portions of some lines. Let's call that fr. 1(b). A third fragment has small portions of the ends of a column and more substantial portions of the beginnings of another column. Let's call that fr. 1(c). Lobel-Page aligns these fragments so that the first line of fr. 1(a)'s col. ii ends with the third line of fr. 1(b), and the last line of fr. 1(b) lines up with l. 1 of fr. 1(c)'s column 2. I see absolutely no reason to even join these fragments together in any way. Lobel-Page attempts at showing a fragment division by writing "fr. 1(a)" at the very top of the combination, and "fr. 1(b)" between l. 21 of fr. 1(a)'s column 2 and the line that lines up with l. 10 of fr. 1(c)'s column 2. This is nowhere near something that matches the actual fragments. To top it all, Campbell says this is a commentary. Since none of the joinings produce lacuna-free lines, I will just follow the actual division of the papyrus fragments, so I'll have fr. 1(a) col. i, fr. 1(a) col. ii, fr. 1(b), fr. 1(c) col. i, and fr. 1(c) col. ii, as separate bits of the same untranslatable thing. Looking further into the numbering, I found this, which quotes part of this thing as "P.Oxy. 2293 fr. 1a col. ii 5 ss." on p. 124. If that numbering is correct, it seems that, for some bizarre reason, my 1(a) and 1(b) were both numbered 1a, and my 1(c) is 1b. I repeat, nothing, NOTHING, in those fragments even remotely suggests a physical join to me, much less that two of them are the same fragment. But I guess this is indeed what was done, because the Lobel-Page collage has fr. 1(a) and fr. 1(b) written in places that somewhat match this bullshit numbering. So I guess my 1(a) will be 1(a1), my 1(b) will be 1(a2), and my 1(c) will be 1(b), because the two 1(a)s are still two and no-one is gonna convince me otherwise. What are we going to do here though? Now, the 1(a)s do join nicely in terms of convincing supplements, so I will put them together as Lobel-Page does. However, the other fragment has no more reason to be in this mix that any of the other P.Oxy. 2293 fragments, as far as I can see, so I will split them. Also, I will split columns apart. fr. 1(a1) col. 1 is thus downgraded to Gibberish class, ending up at 1.A.li. This item only covers# fr. 1(a1) col. ii and fr. 1(a2).

What I gather is that, for some reason, 1(a1) and 1(a2) are both numbered as 1(a), and the other one is 1(b), despite the two not being physically joinable in any way. That is why I have 1(a1) and 1(a2) in the first place, and not 1(a), 1(b), 1(c), as would make sense. That said, 1(a1) and 1(a2) do kind of join nicely in terms of convincing supplements, as the Italian edition mentioned above heavily exploits for its supplements, so in terms of poem division - or rather, text division, because this seems to be a commentary about Sappho - I will take the few letters of fr. 1(a1) col. i on their own, and they will be class 2.A, and then the rest of 1(a1) and 1(a2) will go together with 1(b) col. i in a single class-2.D fragment, and the rest of 1(b) will be another 2.D fragment.

P.Oxy. 1231 fr. 19 + P.Oxy. 2166(a) fr. 4b: two fragments, a handful of letters

Image:

Since the joint line is an isthmus, I will refrain from splitting the transcriptions apart: the last three lines are from P.Oxy. 2166(a) fr. 4b, which gives the beginning of the fourth-to-last line as well, while the part of that line after the ||, as well as the lines above that, are from P.Oxy. 1231 fr. 19. Let's transcribe these then:

]Π€ΠΛ[ ] . . +.+ ΟΡ̣Μ̣ΟΙϹ[.]Τ̤€̣[ ] . . . . . . +. .+ [.] ω̣ [ ] . ΑΛ̣ +. . .+ . [. .]Α̣Π̣ΟΙ̣[ ] +. . . .+ . ω̣[. . . .]Τ̣[ ] . Ι̣Γ̣Ο̣[.]+. .+ . . +.+ . +. . . .+ .[ ] .· ] . [.] ΛΜ̣[.] +. .+ [.] +. . .+ || +.+ [.] . [.] . [ ] . ΙΤ€Ι +.+ . Γ . Ι +. . + . . [ ]Δ€[. . .] . . [ . . ] . [ ] . Μ . [

] πέπλ[ον τ]οὶ+ς+ ὄρ̣μ̣οις [.]τ̤ε̣[ ] . . . . . . +. .+ [.] ω̣ [ ] . αλ̣ +. . .+ . [. .]α̣π̣οι̣[ ] +. . . .+ . ω̣[. . . .]τ̣[ ] . ι̣γ̣ο̣[.]+. .+ . . +.+ . +. . . .+ .[ ] .· ] . [.] λμ̣[.] +. .+ [.] +. . .+ || +.+ [.] . [.] . [ ] . ιτει +.+ . γ . ι +. . + . . [ ]δε[. . .] . . [ . . ] . [ ] . μ . [

Reading notes:

• In l. 2, the second vestige is the bottom of a vertical line, which could be an iota, but also several other letters, hence the vestige status; the omicron seems weird, but it's an effect of pixelation plus the strange lighter line in the papyrus; rho and mu are both half-gone; at the end, we see the rightmost part of a horizontal line and a dot below, which could mean tau or gamma, and I picked tau out of tradition continuity, but gave it two underdots; then there is an arc, which could be an epsilon or sigma, or more round letters, and again tradition continuity determined the choice;
• In l. 3, LP reads «].[...].[.]ω̣»; what I see is: a vestige which is a dot in that "hunch" at the far right of the line, which is probably the first trace he marks; then a little horizontal macron-like trace, some black near the hole, and an arc, which I assume matches LP's «[...]»; then another arc, which matches the next trace; then there is some space, which I think is accidental, and LP seems to treat as a one-letter hole; finally, one last trace, which looks kinda like an omega; then LP says the column ends, but I seem to see some more ink further right, after a two-letter gap which would then be blanked out, and a one-letter hole;
• In l. 4, I see a possible vestige on the far left, perhaps just a high dot; then an alpha; then what seems to be a faint-ass lambda; then a 2- or 3-letter blanked-out space; then a trace; then a two-letter hole; then a mess :); no seriously, I think there is the belly of an alpha, whose slanted right part would hit a vertical leg which is the left portion of a pi, and those are uncertain, then we have a pretty solid omicron, and what could be an iota; so I ended up with LP's text, except he treats the uncertain lambda as non-existent and chucks it with the blanked-out space into a three-letter hole; also, under the pi there is a circle, which should belong to the next line, but LP doesn't seem to have anything of the likes… weird…;
• In l. 5, the tau's vertical leg is barely visible, hence the underdot;
• In l. 6, the reverse-თ-shaped vestige between my last two ++ gaps is not acknowledged by LP, but I can definitely see it;
• In l. 7, I see two dots, so I took the last one for a high dot, and the fainter first one as a remainder of a letter;
• In l. 8, the mu could be a lambda, but that seems less likely because the falling slanted line seems to curve back upward;
• In l. 9, LP sees a nu, but I see a slanted stroke and a vertical one, totally separated, so no nu seems warranted, and instead we have a vestige, possibly an alpha or lambda, and then an iota; at the end, I don't think there's a missing letter between the two vestiges: that space could be an inter-letter space - a kerning, if you want.

P.Oxy. 1787 frr. 7(a) + 7(b) + 19 + LSM IV 9(a) + 9(b) + 9(c1) + 9(c2) + P.Oxy. XXI add. p. 135: a multitude of fragments with way too few numbers and decidedly arguable joints

Image:

Based on "The Oxyrhynchus Papyri" vol. XV's transcriptions of just the P.Oxy. 1787 fragments, the joint lines should be roughly the following, plus possibly more in the extra fragments:

The bracketed parts of the numbers are all my additions. Pardon me for not tracing the joint line of 7(a) and 7(b), but you can clearly see a gap between the two. Anyway, looking at the image, you have 5 groups of fragments that evidently physically join together:

1. 7(a), LSM 9(a), XXI add.;
2. 7(b), LSM 9(b);
3. LSM 9(c1);
4. LSM 9(c2);
5. 19;

Now, the first two go together because 7(b) completes an alpha found in 7(a). The join is also quite snug, so they should be physically joint together, except someone decided to sellotape them when misaligned. WTF? Well, anyways, we still have 4 groups. Groups 3-4 go together on grounds of convincing supplements: a complete word emerges from putting them together. Why the group of 7(a) and that of 9(c1) should join is a mystery. LP gives no reason, and the way he joins them can't be a physical joint because it leaves a 1-letter gap, and with these tatters it's hard to see convincing supplements, so… why? And fr. 19 apparently is "ex eadem parte papyri", from the same part of the papyrus, again, Idk why. So in the end, what comes out are 68(a), which has all but fr. 19, and 69(b), which is 19, and could be joined with the rest with some gap in between whose extent is uncertain. In any case, transcription time!

P.Oxy. 1787 fr. 7(a) ]Ι̣ΓÁΡΜ'ΑΠΥΤΑϹ[ ].́ΜωϹΔ'[ ]ḮϹΑΝΘ€ΟΙϹΙΝ[ ]ΑϹĀΝΑΛÍΤΡΑ[ ]ΔΡΟΜ€ΔΑΝ̣[.].[ ] . Α . . . Κ . [ LSM IV 9(a) ]€ . [ ]€Γ€Ν[ P.Oxy. XXI add. p. 135 ] . Αξ[ ]Α· ] . ÝΝΗ[ ] . € . [ P.Oxy. 1787 fr. 7(b) ] . [. .] . [ ]ΡÓΠΟΝΑ[ ]ΟϹ̣ . ΝΟΥΚΑΤΙϹ[ ] . ΥΝΔΑΡÍΔΑΙ[ ]ΧΑΡÍ€ΝΤ'Ă . [ ]ΗΚ€́ΤΙϹΥΝ[ ]ṆA[·Β·   ·Κ]ΑΗ·[ LSM IV 9(b) ]ΑΡ[ ]€ΟΝΔ€́Τ[ ]Κ[.] . [ LSM IV 9(c1) ]ḲA[ ]ΑϹΥ[.] . . . [ ]Κ'ÁΔΟΛ[ ]ΜΕΓΑΡΑ[ LSM IV 9(c2) ] . [ ]KA[ ]ON[ ] . [ P.Oxy. 1787 fr. 19 ] . . . . Φ[ ] . [.]'ΘÝΡΑ . [ ] . ΟΙΧÁΛ€ . [ ]Δ€ΚÝ[ ] . ΟΠÁΛΗΝÓ . [ ]€[

P.Oxy. 1787 fr. 7(a) κα]ὶ̣ γάρ μ' ἀπὺ τᾶς [ ] ὔμως δ' [ ] ἴσαν θέοισιν ]ασαν ἀλίτραν[ Ἀν]δρομέδαν̣[.].[ ] . α . . . κ . [ LSM IV 9(a) ]ε . [ ] ἔγεν[το P.Oxy. XXI add. p. 135 ] . αξ[ ]α· ] . ύνη[ ] . ε . [ P.Oxy. 1787 fr. 7(b) ] . [. .] . [ τ]ρόπον ἀ[ ]ος̣ . ν οὐκατισ[ ] Τυνδαρίδαι [ ] χαρίεντ' ἀ . [ μ]ηκέτι συν[ ]ν̣ ἀ[κ·β·άλ·κ]αη·[ν LSM IV 9(b) ]αρ[ ]ἔον δὲ τ[ ]κ[.] . [ LSM IV 9(c1) ]κ̣α[ ]ασυ[.] . . . [ ]κ' ἄδολ[ ] Μεγάρα [ LSM IV 9(c2) ] . [ ]κα[ ]ον[ ] . [ P.Oxy. 1787 fr. 19 ] . . . . φ[ ] . [.]' θύρα . [ ] .οι χαλέπ[ ]δεκύ[ ] . οπάλην ὀ . [ ]ε[

Reading notes:

• In l. 5 of fr. 7(a), the final nu of Andromedan is half-gone and could be a mu, hence the downgrading to underdot.
• In l. 6 of fr. 7(a), I see: a horizontal line, which only allows a tau or a gamma since no vertical lines or slanted lines start from it; an alpha, which is in fact completed by 7(b); three dots, which I count as three vestiges, but the first two might be the tops of a mu; what is undoubtedly the top of a kappa, with the top slanted stroke not reaching the top of the letter and this observation excluding all other letters; and a ^-shape, possibly an alpha.
• The first vestige of l. 1 of the add. fragment is part of the same letter as the last vestige of l. 5 of fr. 7(a).
• The final alpha of l. 2 of this fragment looks a lot like a chi, but I'm quite persuaded that there are papyrus quirks at hand; however, part of the wannabe quirks looks really like ink, whence the high dot in the raw transcription.
• L. 1 of fr. 7(b) starts with a dot that is part of the alpha in the last line of fr. 7(a), then the space of the three dots of that line is traceless, then there is a dot which is probably part of our kappa's bottom, or it might line up with the last of the famous three dots.
• In l. 3 of fr. 7(b), the vestige is basically a dot, little more; the uncertain sigma is revealed by LSM IV 9(b) to actually be a rho, of which we only see the tail, but when we get to see an extra arc detached from the body, we must conclude it's a belly; so essentially, C̣ + arc = P is our addition.
• In l. 4 the vestige is presumably part of a horizontal line, so we could have a gamma, not a zeta because we'd see a slanted line starting from top-right and that's not there, a pi, a tau.
• In l. 3 of fr. 19, the first vestige could be a sigma, a mu, or possibly other letters, hence the downgrading from uncertain mu.
• The last vestige of l. 5 can be an iota, a nu, or other options, hence the downgrading from underdot nu.

And now let me collage all these minus fr. 19, which is separated from the others. 7(a) is black, 9(a) blue, add. red, 7(b) green, 9(b) purple, 9(c1) brown, 9(c2) pink -- good luck colorblind readers :).

κα]ὶ̣ γάρ μ' ἀπὺ τᾶς ἔμ[ας
] ὔμως δ' ἔγεν[το
] ἴσαν θέοισιν
]ασαν ἀλίτραν[
Ἀν]δρομέδαν̣[.].[.] . αξ[
]αρ[. . .] . α . . . κ . [.]α·
]ἔον δὲ τρόπον ἀ[.] . ύνη[
]κορον οὐκατισ[.] . ε . [
]κ̣α[. . . . .] . Τυνδαρίδαι [
]ασυ[.] . . . κα[ὶ] χαρίεντ' ἀ . [
] κ' ἄδολον[μ]ηκέτι συν[
] Μεγάρα . [. .]ν̣ ἀ[κ·β·άλ·κ]αη·[ν

P.Berol. 9722 fol. 1: another crazy-faint parchment with wildly varying readings

Image:



Like the last P.Berol. 9722, here is Lobel-Page vs. Edmonds:

Again, the variation is pretty wild. This hasn't been fully transcribed yet, but I wanted to start the section. I can tell you it looks like Edmonds switched ll. 1-2 with one another, however. I mean, his l. 1 looks like l. 2, and l. 1 in the parchment could possibly be found in l. 2 of Edmonds' text. I haven't gone further, except I know l. 5 in Edmonds and the parchment both match l. 5 in LP.
I say it here and nowhere else: I won't be transcribing fol. 5 verso, because Edmonds doesn't have it, and I'm sick of destroying my eyes over these damn faint-ass parchments, so I'll just trust LP for that one.
That said, I handled this in much the same way as fol. 3 above, except instead of manually imitating the shape of the signs, I retraced them directly on the image. So here is the image with the retracing done. Black means definitely there, blue means quite visible but slightly faint, red means visible enough, yellow means so faint it's hardly visible (what was faint red in fol. 3). Areas surrounded by thin green lines are those where I flat-out refused to try to retrace anything.



But it doesn't end here. See that half-done line? Well, while retracing it, I realized stupid app Preview had blurred my image, as if retracing this on high definition wasn't hard enough, so I cropped out the rest from the high resolution original and went on as follows.



But this was not the end either, because, while looking randomly around, I had noticed signs in an area of the parchment where I had no transcriptions at all. Oh and also, the last line on the left is also nowhere found in transcriptions. Bottom-right, zoom in, and retrace:



Some of those letters are quite convincing, and given the great lengths I've gone to in the rest of the parchment, it really seems absurd not to transcribe this as well. There are also a couple easily guessable joint lines I will take into account. Precisely, l. 19 will have two || to mark the lines, ll. 20-21 will have one ||, and that's it. That's because in l. 19 is where the point of that upward-V fragment goes in, so the first || enters said point, and the second || leaves the V and goes into the separated mini-scrap near the big middle hole, while the other two lines have a piece on the V and a piece on the mini-scrap, then the mini-scrap ends and we're always on that V. Now it's time to give a bunch of decryption notes to figure out how to transcribe that mess, and then give a colorful raw transcription and the final text. But that will take ages, so have patience. For the moment, I have the first 11 lines.

1. In l. 1:
   • The beginning is clearly an omega, which will average to blue;
   • The following is a rather convincing phi, averaging to red;
   • Then there are very thin apparent alpha-iota-nu-nu which are entirely yellow;
   • It seems the scribe was having fun or maybe that apparent depiction of male genitalia between the first two lines is a papyrus quirk :);
   • The following scribble I take to be blue alpha + blue mu (which ends at the black) + blue papyrus quirk + another mu averaging to blue + yellow quirky apparent connection to the following letter + blue actual connection to it + red iota + yellow quirks + kappa averaging to blue because most of it is blue and that part is enough to decide it's a kappa;
   • The hole is one letter;
   • Then we have an eta averaging to red, connected in blue (with underside quirks) to some half-red-half-blue letter which looks like an epsilon that's mostly red;
   • Then I can't tell what is quirks and what is signs, so I'll put a vestige in blue, which looks like a chi but again, quirking through it I can get to an upsilon as in Edmonds' text;
   • Then there is a clear nu with one very thin red leg and most of the rest black, which makes it black since that shape couldn't be attributed to another letter AFAICT;
   • Then there is an alpha which will average to blue;
   • Then a yellow mess which can be construed, if you squint hard enough, as a small nu between the alpha before and a lambda, the nu sitting slightly higher than its neighbors; so it will be nu+lambda, both uncertain and yellow;
   • Oh btw, uncertainties:
     • The first two letters are certain;
     • The AINN are underdot because of the thinness;
     • The AMMI is all uncertain because it's a messy scribble;
     • KH certain;
     • Epsilon underdot;
     • Then we said vestige;
     • Then we have certain NA, and we're back to our mess;
   • After said mess, the upsilon is visible, and averages to blue because the bottom tells it from a theta or omicron and is blue;
   • The rest is fairly straightforward, requiring only to say that the gamma might be a tau but that seems less likely given the right and left extents of the horizontal top, and that from that first red mu on I wager all is but papyrus quirks; oh and also, right after the upsilon only the two slants (yellow and red) count as l. 1, what is below will be quirked away, except the black which will be seen as a continuation of the red slant;
2. In l. 2 we have:
   • Vestige at the start for that apparent U;
   • Then the psi is a bit messed up but certain enough, averages to red;
   • The following is apparently a lambda, but I'll read a double-underdot black alpha since psi-lambda isn't a thing;
   • Then we have a black pi;
   • Then I'll have to be somewhat bold and ignore the blue stroke as a quirk but take the red circle into account to form an average-blue phi which is weirdly shaped and thus earns an underdot;
   • The red leg, yellow slant, and black reverse slant form a red-average uncertain eta;
   • The rest of what's before the hole is a red-average mu -- red because most of it is yellow so we give it the average faintness closer to that part --;
   • The hole probably harbored no letters;
   • What follows seems to be a blue overlined mu;
   • Then a red alpha, both certain;
   • Then two vestiges, one blue and one yellow, with the infinity above being treated as a quirk;
   • Then a red-average omicron;
   • Then +.+ for the greened area;
   • Then a blue vestige;
   • A blue-average pi (legs black, top part-yellow-part-blue);
   • The blue line slanting out of the pi is a quirk, and the two slashes, one yellow and one red, give a yellow vestige;
   • Then a blue phi;
   • Then we seem to have a pi, but phi-pi isn't a thing (at most the other way around), so I'll go with a blue and a yellow quirk, and a black iota;
   • Then it seems I have a black gamma with a red quirk;
   • Then an average-blue epsilon, blue because the blue top makes sure it isn't the weirdest sigma ever;
   • Then I'd guess lambda-alpha-nu, but all shapes are weird, and the first two are almost identical; could we read the last thing as an omega? Very unlikely… I think I'll make it three black vestiges;
3. In l. 3, we see:
   • A mess to the left of the supposed line beginning which I'll take as a quirk of various colours;
   • A pi with red top and blue legs, which averages to red because without the top it could be an eta;
   • What looks like an epsilon of various colors, mostly red, hence the red average, and I say "looks like" because it has one stroke too many (it's basically a euro sign €), so underdot;
   • An almost-entirely-red phi which is certain;
   • What looks like a yellow alpha but deviates from the usual shape of alphas in this thing, so underdot;
   • What seems to be +.+;
   • A yellow nu;
   • What looks like a half-black-half-blue iota with a margin quirk below it; yes, the quirk is black, but the margins often darken here, so that black can be the result of darkening; also, the iota averages to black because even without the bottom it would have been unmistakeable;
   • What is either a half-nu or a half-mu, and could be recognized even without that blue leg, so averages to black;
   • A hole of 2 or 3 letters; say 3, let's be generous;
   • What is possibly part of a blue rho, underdot for you;
   • What seems to be a nu with a half-blue-half-black left leg and a red rest, average red;
   • A thin yellow nu with black bottom dots, I'll make it a black vestige because I'm always skeptical of thin letters;
   • A yellow epsilon with some black quirks, or maybe a part-black-part-yellow theta or omicron, I'll make it a yellow vestige;
   • A blue vestige which is basically an underscore; in fact, scratch that: it's an acute accent in the next line, so we have +.+;
   • A blue chi with a red quirk, followed by +.+;
   • A half-black-half-blue nu which averages to blue because the black half could be construed as a half-omicron or such;
   • A red upsilon with red acute accent;
   • A blue delta;
   • Yellow and blue traces suggesting an alpha, which is mostly yellow, so in the end, yellow underdot alpha;
   • Either a slanted nu or a weird tau (in the style of the ostrakon), so black vestige (because in both cases it's black);
   • A blue top-of-upsilon which could be construed as a bottom-of-theta or such, hence double underdot;
   • An overlined blue delta (why the overline? let's say quirk :) );
   • A messy mostly-blue epsilon which gets an underdot;
   • A half-blue-half-black delta which averages to blue because its bulk is that color;
   • A blue iota;
   • A red delta;
   • A mostly-yellow epsilon;
4. In l. 4, we find:
   • A strange red kappa with what is probably a blue quirk, possibly part of the mess just above;
   • A yellow sideways T which is a quirk inserted inside what is probably a half-blue-half-black rho, which I'll average to blue with underdot;
   • Weird quirky connections of that to a red iota;
   • A very colorful nu which averages to red;
   • Something that, however you look at it, is either an omicron or a chi, but I'll read as blue-average double-underdot omicron because this way I get a word out of all this and the next letter;
   • A half-yellow-half-black nu which averages to blue because most of it is black;
   • Two black vestiges;
   • A 4-letter hole;
   • A half-blue-half-yellow vestige which I average to red;
   • Something that looks like a strange blue glyph ᛗ (futhorc letter "man") from an exotic script but is probably a weird mu, so underdot;
   • A blue eta with the famous acute accent which has been scrapped from the previous line;
   • A red round quirk, and what is probably a half-blue-half-black rho (the horizontal thing doesn't go into it; average blue, underdot for "probably";
   • The apparent blue descender is probably an overgrown ascender from the following line;
   • A… nu? More like, a mess; blue vestige, let me say, because it's part-black-part-red and there's even some blue;
   • The continuation of the mess, which I wager gives a yellow vestige and a part-blue-part-red double-underdot red-average epsilon;
   • A yellow pi sitting on top of a red quirk;
   • A blue-average epsilon which is certainly an epsilon thanks to the black middle, but Edmonds probably took as an alpha due to the red stuff, which is probably quirks;
   • How on earth is that thing a rho? More like OIA, all blue or average-blue; though that overgrown iota has me perplexed;
   • A mess which I'd say gives a yellow and a red vestige;
   • A yellow alpha;
   • A red iota, connected in blue and yellow to the alpha (two connections);
5. In l. 5, we start seeing big messes:
   • We start with a yellow quirk;
   • Then a black-average pi which only has a little bit of the bottom-left blue and is black-connected to the following letter;
   • Which is a blue-average epsilon, mostly black but the red bottom is what distinguishes it from a sigma, hence the average;
   • Then a red pi;
   • Then a half-black-half-blue lambda, with a blue dot in the middle which makes it look like an alpha (so underdot), and a blue wannabe correction which I'll dismiss as a quirk or error because I can form a full word with the lambda;
   • What next? Do we read ON and then asterisk? I think I'll go for an omicron formed by the blue outer things (the inner one being a quirk as well as the yellow one) and the red bottom;
   • Then yellow quirks forming a vertical leg and an asterisk;
   • Then a blue weird nu which gains an underdot; the omicron above also deserves one;
   • Then a vestige I average to red, it's red and yellow + a black "dot";
   • Then a black iota, where the black top is ascribed to the margin darkening;
   • Then what would seem to be a black phi, but could be a psi, so underdot;
   • The hole doesn't affect this line because it's at its top, but no sign is visible for 2 letters's space, so +. .+;
   • Blue strokes forming an apparent lambda which is squished and is actually probably a chi, so blue chi be it;
   • At this point that black-blue-red contraption can't be a chi (chi+chi isn't a thing in Greek AFAIK), so blue vestige; in fact, treat the red as a quirk, and you get a black-average nu;
   • What follows is a bloody mess where I can hardly see anything; we have that blue-red-yellow contraption which is definitely going to be a vestige, I'll say blue-average because it's almost entirely that color; then what would seem to be a black iota but is too short to be full and could then be an eta, hence iota-underdot; then I'll treat that yellow nu as a quirk and see another black-average iota, at which point I wonder if they formed a mu or nu together, so underdot again; then two red ripples and a blue line which will have to be a red vestige;
   • Then finally a clear letter: mostly yellow, little bit of red, so yellow-average gamma;
   • What follows would seem to be a sigma, but gamma+sigma isn't a thing in Greek AFAIK, so I leave it as a black vestige;
   • Then a blue omicron;
   • And after a 2-letter hole, IA, i.e. black iota and blue-average alpha, blue because the middle is red and is what differentiates it from a lambda;
6. In l. 6, more messes:
   • KAI is clear enough, though that alpha is strange; red-average kappa (it has some yellow but just the red would be enough to recognize it), red-average alpha (it has all colors, but most of it is yellow, and half of it is controlled by yellow and red, the left half), and finally blue iota, because the black part is little;
   • Then a weird kappa which I'd take for a chi, so underdot, and black-average, because it's mostly black, and would be recognizable without the blue;
   • The blue partial infinity is a quirk to me;
   • Then a lambda, half blue half black, so blue-average;
   • Then possibly the weirdest epsilon ever, underdot for weirdness, blue average because it has red and blue in the middle and black at the endpoints, and in particular the famous horizontal middle (which is slanted here) is blue;
   • Then a blue iota;
   • Then I'll assume the yellow is a quirk, otherwise I'd end up with three sigmas which is not possible, and so the red produces a iota, party faded into yellow and partly red, connected to the following letter in red;
   • Then two sigmas, one half-blue-half-black so blue-average, one almost all black and recognizable even without the blue, so black average;
   • Then a half-blue-half-black alpha with the left half being blue and winning the average;
   • Then a strangeish omega which has black sides and red middle, so blue-average;
   • Then a red chi, connected in yellow to the following letter;
   • Which is a part-yellow-part-black nu I would probably recognize without the yellow, so black-average;
   • Then a 1-letter space which I'll mark as faded +.+;
   • Then a black lambda;
   • That red chi is squished between two letters and also follows a lambda, which is impossible, so I treat it as a quirk;
   • An omicron follows, left-red-right-blue, right is enough to identify it IMO, so blue-average;
   • Black vestige (a dot basically);
   • ICA, i.e. half-red-half-yellow iota which I average to red, red sigma, left-yellow-right-red alpha which averages to yellow;
   • Then a two-letter space which I should probably have greened out because I see signs, but let's say +. . +;
   • Then what looks like a black omega, a bit strange so underdot, connected in black to the previous letter;
   • And finally a black sigma;
7. In l. 7:
   • A rather unescapable kappa which averages to blue;
   • A clear enough rho which averages to blue since it's half-red half-black;
   • A little further right, another rather unescapable kappa which averages to blue since it's 3/4-black 1/4-red;
   • In between, a mess of black, yellow, and red, with even a breve in blue which is either a quirk or from the previous line; I will take the leftmost little yellow blot as a quirk, the rightmost black as mostly a ligature, and the rest as part of an underdotted red-average omicron;
   • After that kappa, the omicron, averaging to blue since it's 2/3-blue 1/3-black, is unescapable;
   • Linked to it, an epsilon which averages to blue since, despite being mostly black, the red part is the one that distinguishes it from a sigma (the mythical middle stroke we have oft used to disambiguate :) );
   • The black nu is clear, continues in red into a quirk, and the X above it in red is a quirk, so this is a black nu;
   • Lots of quirks here, but the tau I see, which is underdotted because it requires some serious squinting to be seen, is formed by some blue and some red for the top, and the only black stroke for the bottom;
   • THe following alpha is much clearer, if oddly-shaped; average blue since that's the biggest contribution; the yellow is a quirk;
   • The two blacks and some of the yellow form a lambda, and the rest of the yellow is quirks;
   • What follows is an omega with a rather odd shape, and since the oddity is in black and most of the omega is in blue, it'll have to be blue-average and underdotted;
   • The following black curved line seems to be a half-omicron, the rest undistinguishable from margin darkening;
   • What follows surely seems like a rho;
   • Then we have a red-average epsilon with an extra yellow quirk on top;
   • Then I'll have a yellow-average vestige for all the rest until…
   • …the clear blue sigma and the clear red-average (red disambiguates from sigma once more) epsilon;
   • The greened-out area is a single letter;
   • BA is then clear, and the alpha connects to the following letter. all in blue;
   • Then we have a black lambda, a blue tau which might be a quirk so double underdot, a half-blue half-black blue-average eta, a black K with a blue link to the following black omicron, which in turn black-links to the following half-black half-red blue-average iota.
8. In l. 8:
   • That pi is a challenge; black bottom, top starting blue, dipping down yellow, and ending red into the following letter; blue-average since red and yellow are small contributions and the bulk (aka bottom) is black;
   • The next letter is a clear red epsilon;
   • Then there is an odd pi: top starts blue and curves so much it seems to join with the blue left leg, so top starts blue, continues red, ends yellow and links to the following lambda, whereas the legs are one blue and one yellow; red-average, and underdot since it seems to be sigma-tau;
   • Lambda is blue-average since most of it is blue; the semicircle above its link to the following letter is a quirk;
   • The following letter is a small half-black half-blue blue-average omicron;
   • The following mess is probably a nu, where the left leg is black, the slanted stroke is blue, the right leg is mostly red, and both legs are so thick and disuniformly faded that they seem to be omicrons or similar; blue-average and underdot;
   • The following ΠΟΡΦ is sufficiently clear, pi being blue-average and having contributions from black to red, omicron being half-blue half-red blue-average and seeming to carry a circumflex which is probably a quirk since it makes no sense, the rho is somewhat oddly shaped, especially for this papyrus, so it gets an underdot, and is blue-average since it has black and red contributions, and the red on top is a quirk, and the phi is clear, having contributinos of all color but getting blue average;
   • Then we have what definitely seems to be part of an Y, mostly black partly blue, I'll say black average and underdot;
   • Hole is one letter, then ION is clear, iota being half-red half-blue blue-average (the red part is way up top and could just as well be treated as a quirk), the other two letters are yellow;
   • The following mess accounts for two letters, perhaps with a one-letter gap in between;
   • Then we see a red epsilon, a half-blue half-black nu with two red dots which averages to blue and is certain, then a half-blue half-red bottom-left of delta which averages to red because the red distinguishes it from a lambda, the hole contained the rest of it;
   • Then we have a 2/3-black 1/3-blue epsilon averaged to black because the middle stroke is, then a black xi which I wager could be a zeta so underdot, the what would seem to be XO but the chi-xi cluser isn't a thing in Greek so I'll have to squint and see an omega with quirks there, underdotted and blue-average;
   • The line ends with a greened-out area of about 7-8 letters.
9. In l. 9:
   • Squinting hard enough at the beginning mess, I can see a chi which is part-red and part-yellow, gets two underdots for bare legibility, and a red average, and then there's a bunch of quirks which hinder the legibility of the chi; then there's a half-red half-black short lambda with a big yellow quirk to the top-right;
   • At this point what follows cannot be an actual phi since chi-lambda-phi is not a thing, so that big yellow thing is going to have to be a quirk, just like the red thing below it;
   • What follows is a clear alpha, which averages to yellow since most of it is yellow;
   • That apparent R is, IMO, actually a half-yellow half-black red-averag iota which was traced bottom-up and linked to the first leg of the following nu which came out squiggly;
   • Said nu has the squiggly yellow leg, the red slanted stroke and the blue right leg leading into the following letter, so average red;
   • The following letter appears to be an omicron, but IIRC I looked at the non-blurred image again and the round thing connects to the black bottom-of-iota, making this a red-average alpha;
   • Then we have a half-yellow half-black red-average iota;
   • The following looks like a nu, but everyone says pi… let's see the non-blurred image again… OK so what actually happens in that omicron/alpha is that the black stroke continues downward, not reaching the other black thing… as for this nu/pi, atop what I have there is a super-faint yellow horizontal top, which makes this definitely a pi, and the red in between is probably the legs thickening;
   • The following epsilon is clear, the \-blue is a quirk, then we have a blue rho tied to the following 3/4-blue 1/4-black blue-average phi, then a hole of 1-2 letters;
   • The blue and black mess that follows can be read as either black OII, blue rho and black iota, or half-blue half-black blue-average phi plus two iotas, but two iotas is a nasty combination, so I'm going with the rho-iota reading;
   • Then we have a black sigma with a red quirk which is justifiable as the bottom thickening; the theta is assigned to the next line;
   • Then we have an odd 2/3-red 1/3-black red-average underdot alpha;
   • The two blue strokes presumably form a nu with a stroke lost to the hole, could be a mu so underdot, obviously blue average;
   • The following black omega is clear enough;
   • Then we have a messy nu with contributions from black, blue, and yellow which gets a blue average since most of it is blue;
   • What follows is a mess of all colors, with red being the biggest contributor, so a red vestige;
   • Then we have a nu with a thick slanted stroke which is mostly blue, so blue average, but has contributions from all colors;
   • Finally, a greened-out area of 6-7 letters.
10. In l. 10:
    • The first two strokes are from the line below, as we shall see; then there is a half-blue half-black blue-average sigma which seems to tie to something before it and leads into the following letter;
    • That letter is a mostly-black black-average tau with a little yellow part, which ties into the following letter;
    • That letter is doubled: there is a black version on the bottom with a blue round quirk inside it, and a blue version on top, presumably a reading aid added by a second scribe;
    • What follows is the weirdest quirkiest phi ever, with black yellow and blue contributions, but mostly blue, so blue-average and double-underdot;
    • Then there is a clear enough red alpha;
    • The two strokes are probably a nu which lost its middle to fading, one is black and one red so blue-average, could be a mini-mu or an eta so underdot;
    • The hole contains one letter, and then we have a half-red half-yellow mostly-yellow yellow-average iota;
    • Then the pi is rather clear, top is black, bottom is red, average blue;
    • The pi ties to the following rather messy epsilon, which has a small black contribution, some yellow, and mostly is red; in fact, the black and part of the yellow are probably quirks, so red-average; since it's so messy, underdot;
    • That triangle is probably what's left of a rho, but definitely underdotted, and since it's 2/3-black and 1/3-red, the average is blue;
    • The hole starts getting big; 11-12 letters in this line;
    • Then there is a half-omicron which, IIRC, I saw competed in the non-blurred image, I'll say blue-average based on what I have in this image;
    • The following mess is probably a nu which has a black part and a blue right leg leading into the following letter, and then a blue quirk and a red one, the latter leading into an epsilon; I'd read iota-epsilon there, and then a 1/3-blue 2/3-red red-average vestige; let's see what Edmonds has to say…
    • No way I can see δέθεντες, but if I squint hard enough on the non-blurred image I can see that the wannabe omicron has a black thing in it, so it's a blue-average theta, then the wannabe nu is actually a quirky blue-average epsilon, and then there are a red-average nu and the top of something, maybe a tau, let's make that underdot and red, and then nothing decently certain;
11. For l. 11, I re-retraced the beginning because the blurring had destroyed my ability to retrace it:
    
    the end seems to have a iota and a phi, let me see…yep, blue iota and red-average… actually, alpha, with contributions from all colors except black; also, the hole is 11-13 letters;
    • I'm gonna be cautious and dismiss that initial omega as a quirk; then we have a kappa which has a red leg, a yellow top-right, a blue bottom-right, and has lost the middle; the blue ties it with the next letter;
    • With some ties and some quirks, it's not too hard to construe the following yellow and black mess as an alpha; underdot for construing, and yellow-average since most of it is yellow;
    • With some thickness arguments, we can easily get a half-blue half-black lambda next, with lots of quirks above it, and a thick yellow one below it; black-average;
    • The half-blue half-yellow epsilon on top is probably a quirk; the following mess will be treated as a blue-average vestige, but it does suggest a theta;
    • The we have a half-black half-red epsilon/omicron/theta, so a blue-average vestige;
    • The following is a black sigma with a red quirk, which is dismissed because it's thicker than the rest;
    • Further right we have a half-red half-blue quirky alpha, red average and underdot for quirks; in between, one would say omicron, but OCOA isn't a thing AFAIK, so I'll dismiss the red as a quirk and treat the rest as a half-blue half-black blue-average square sigma;
    • After the alpha, a tying red mu with a doppelgänger below it, which leads into a mostly red alpha which has a small blue contribution but definitely averages to red;
    • Finally, a greened-out area of 2-3 letters leading up to the hole.

And this allows me to transcribe those lines as follows.

ωΦΑ̣Ι̣Ν̣Ν̣Α̣Μ̣Μ̣Ι̣Κ[.]Η€̣ . ΝΑΝ̤Λ̤ΥΙΙ€€Γ€ΟΝΑ
. ΨΑ̤ΠΦ̣Η̣Μ[]ΜΑ . . Ο+.+ . Π . ΦΙΓ€ . . .
Π€ΦΑ̣+.+ΝΙΝ̣[. . .]Ρ̣Ν . . +.+Χ+.+ΝÝΔΑ̣ . Υ̤Δ̄€̣ΔΙΔ€
ΚΡ̣ΙΝΟ̤Ν . . [. . . .] . Μ̣H́Ρ̣ . . €̤Π€ΟΙΑ . . ΑΙ
Π€ΠΛ̣ΟΝ̣ . ΙΦ̣+. .+ΧΝ . Ι̣Ι̣ . Γ . Ο[. .]IA
ΚΑ̣ΙΚ̣Λ€̣IΙϹϹΑω̣ΧΝ+.+ΛΟ . ΙϹΑ+. .+ω̣Ϲ
ΚΡΟ̣ΚΟ€ΝΤ̣AΛω̣[]Ο̣Ρ€.Ϲ€+.+ΒΑΛΤ̤ΗΚΟΙ
Π€Π̣ΛΟΝ̣ΠΟΡ̣ΦΥ̣[.]ΙΟΝ+.+ . . €ΝΔ€Ξ̣ω̣+. . . . . . . .+
Χ̤ΛΑΙΝΑΙΠ€ΡΦ[. .]ΡΙϹΑ̣Ν̣ωΝ . Ν+. . . . . . .+
ϹΤ€€Φ̤ΑΝ̣[.]ΙΠ€̣Ρ̣[. . . . . . . . . . . .]Θ€ΝΤ̣ . .
{ω̣}ΚΑ̣Λ . . ϹϹ̣Α̣ΜΑ+. . .+[. . . . . . . . . . . .]ΙΑ


Now we move to part 2, i.e. ll. 12-20.

1. In l. 12:
   • Welp, the first red-and-yellow thing is gonna have to be a quirk, because it could only be a quirky lambda or at most chi, and then there's a wannabe phi;
   • What follows is a blue phi with a blue quirk turning it into a female sign;
   • The next thing is a mess of all colors, where the most likely interpretation is that it's all quirks except the red mu, but phi-mu isn't a thing; the yellow thing does look like a lambda, but LP and Edmonds give a rho… the best match is that the red is indeed ink, but only the legs, and that forms a red rho, but wow, that definitely gets an underdot for magicky quirk explanations;
   • What follows is a way more visible upsilon, average red because it's part-yellow part-blue;
   • Then I'd say the blue top and red bottom form a certain red-average sigma, while the rest of the red is quirks;
   • Then the sigma that follows is clear and blue, with a quirk dot in the middle;
   • What follows would look like a blue epsilon, except it's not lunate as all the others, so it'll have to be a blue vestige instead;
   • The following nu is clear, as is the part-black part-blue part-red omega which averages to blue;
   • Then we have a clear pi with black right leg and red rest, average blue, and finally a rho with left leg half-black half-yellow and right leg black, average blue since it's mostly black:
   • I wager the hole contains 23-24 letters;
   • At the other end of the hole, all I can see is a few black signs, maybe a kappa or upsilon, I'll go with uncertain kappa; very far from all the stuff Edmonds reads there; I wonder if he read vestiges before that? I definitely won't, they'd be too chopped off to be read sensibly.
2. In l. 13:
   • We start out with a yellow pi underlined in blue – the red tau before it is either a correction to the following line or a quirk, and since there is a clear tau in said line, it's a quirk; can't be a letter since it's in between lines;
   • Next letter has contributions of all colors, but definitely looks more like an alpha to me than an omicron; red average;
   • Then we have a greened-out one-letter area followed by a half-black half-blue blue-average vestige, perhaps a rho;
   • Then there is a mess of all colors except black, definitely a red-average vestige (cen't average to blue when there's no black, even if the overwhelmingly biggest contribution is blue);
   • Then it would seem we have a theta that's a little high and averages to blue from it having all colors outside yellow; this is above a greened-out one-letter area; I guess I'll have the theta as a correction to the greened-out area;
   • The red thing that follows is probably a letter tie, going into the following nu with blue legs and yellow middle, average red;
   • This ties in yellow to what is probably an omega that starts blue and ends in a red wave, underdot and blue average since it's mostly blue (and has no yellow);
   • The following nu-pi-omicron is clear enough, if a bit weird since the omicron overlaps the pi; the nu is yellow except for the blue left leg, so red average; the pi has red left leg and yellow rest, so yellow-average; the omicron is partly chopped off, all yellow, and I'd say undoubted;
   • The hole is 18-18 letters wide IMO;
   • Right after the hole, the yellow thingy is probably a tie going into the red thing after it;
   • Said thing is, I believe, a red eta, which has lost its right leg which (almost) overlapped with the following black letter, and thus gets an underdot;
   • Then we have a black thing which seems to be a sigma with funny appendages, possibly quirks;
   • The line ends with a mess of all colors which is either one or two vestiges averaging to blue (most is black).
3. In l. 14:
   • I wager there's a black tau with lots of quirks to the side which seem to form an underlined epsilon with two middle strokes, which probably explains Edmonds' reading;
   • At this point, the red line is a quirk, or I get TK, and the rest if the following wannabe K is a top-black bottom-blue black-average upsilon;
   • The red hooking line on top is probably a quirk, whereas the blue nu below it is not;
   • Then red and black cooperate to give a delta or alpha that averages to blue; I say alpha because of the following;
   • It seems the following mess is two kappas, one having contributions from all colors except yellow, and being tied in blue to the other one, which is mostly blue with a little black; first one averages to blue, second one to blue as well;
   • Then we have a greened-out area which probably extends too far right, and if reduces is 1-2 letters;
   • Then we have a pretty certain top-blue big-middle-yellow bottom-black iota with red average, and a top-red bottom-yellow yellow-average pi;
   • Then a one-letter space, then a 6-7–letter hole, then a 4-letter space;
   • The next thing is probably a top-yellow bottom-red epsilon averaging to the color of the famous middle stroke, that is yellow;
   • Then a clear kappa and nu, both yellow, the nu tying into a round yellow arc which could be an omicron; underdot.
4. In l. 15:
   • A greened-out 4-letter oval;
   • What looks like a modern lowercase alpha with blue top and red bottom, but can obviously not be, so I can only suggest a blue nu with red quirk below it;
   • Then a nu with such a weird shape I must give it an underdot, mostly blue, some red, but blue average;
   • Then a red left semicircle, I'd guess a sigma or omicron, let's say omicron-underdot (nu-nu-sigma isn't a thing, I don't think);
   • Then a messed-up illustration of Thales's theorem from high school :); seriously speaking, either an underlined alpha or a delta with middle stroke, but omciron-alpha sounds nasty, so I'll go with delta-underdot, averaging to blue since there's much black and much yellow and some blue;
   • Then a mess of blue and red I'm not sure I can dismiss as a tie, so I'll make it a red-average vestige;
   • Finally a pretty clear mostly-red partly-black blue-average mu leading into the 6-letter hole;
   • On the other side of the hole, a mere 7-letter greened-out area.
5. In l. 16, we have a single lonely letter before an 8-letter green, a 4-letter hole, a 5-letter green, and possibly more space to the right; LP says certain pi, Edmonds says «mu and I don't bother with critical notation», I say WTF; it definitely has a black vertical leg on the right; there is a half-blue half-red vertical leg on the left; the middle is a mess; we have a blue stroke at middle-hight, which would make it an eta; a yellow one on top, to give the pi; and a yellow downward semi-oval, which could make it a mu; so I'll make it a blue-average vestige;
6. In l. 17, we have a 2-letter hole, a 3-letter green, a 4-letter hole, a 3-letter space; on the left of that:
   • A blue gamma tying to nothing on its left;
   • A red nu below the gamma which is a quirk;
   • A yellow upsilon with weird shape, so underdot, and also, it has blue and black quirks near;
   • A red nu with two red "thick dots" above it;
   • What would seem to be a lambda but obviously can't be so it'll have to be an underdot-alpha with red average since what is left is half-blue half-yellow;
   • And a blu iota; I don't form a nu with alpha+iota because the slant of what would be the left leg of the nu is too big IMO;
   and to the right:
   • An eta which has the left leg and the top half of the right one black and the rest yellow, so average blue, which ties in yellow to the following;
   • A red delta tying in red to the following;
   • A pointy part-red part-blue mostly-red red-average omicron from which a blue nu depars;
   • A part-red part-blue red-average epsilon blue-tied to the nu;
   • An overlined blue mu;
   • A blue omicron;
   • A half-blue half-yellow red-average vestige;
   • A blue phi.
7. In l. 18:
   • A clear black delta;
   • That thing with a red top and a blue bottom, which I could actually read as an alpha-underdot if I assume the diagonal stroke went further up-left;
   • A red sigma which was apparently corrected with a blue-average epsilon which has blue horizontal strokes and yellow round part;
   • NI in black;
   • A 5-letter space I didn't green out for some reason, then a 5-letter hole;
   • What is that shit? Yellow sigma-epsilon overlayed with red upsilon, space, red mess, colorful red-average mess, two phis one yellow one black, blue-and-black mess, yellow pi? Let's see Edmonds one sec…Welp, that's one big locus desperatus to me, starting +.+, then what I just said, and then I could go on with blanks.
8. In l. 19:
   • 11-letter green, 4-letter hole, 2-letter green;
   • I'll say weird tau, blue-average since it's mostly blue with a black contrib, underdot for weird shape and for red quirks;
   • Blue added alpha with red overline;
   • Mostly-blue blue-average epsilon with a little red;
   • Weird-thus-underdot mu with red left leg and yellow rest, blue-tying into nothing;
   • Half-blue half-black blue-average alpha;
   • 3-letter green;
   • Weird-thus-underdot black omega with endpoints prolonged in red;
   • Black sigma;
   • 3-letter green;
   • Probably a half-blue half-red red-average delta with underdot for rotation.
9. In l. 20:
   • 9-letter green, 6-letter hole, 2-letter green;
   • Colorful pi missing only black, red-average, with red circle under it;
   • Blue-and-black mess which I'll treat as a mostly-black black-average omicron with blue quirk below;
   • Two taus, one half-black half-red blue-average, one yellow;
   • Possibly an alpha, underdot, half-blue half-red, red-average because the top is red;
   • Blue tau, half-blue half-yellow omicron tying to yellow pi with top prolonged in black; then, sme space.

And that ends the 20 lines of Edmonds. But there's two more in this pic, and three more in the third one. But first, let's see the resulting transcription.

ΦΡ̣ΥϹϹ . ΝωΠΡ[ 23-24 letters]Κ̣
ΠΑ̣+.+ . . +.+θΝω̣ΝΠΟ[ 18-19 letters ]Η̣Ϲ . .
ΤΥΝΑ̣ΚΚ+. .+ΙΠ+.+[ 6-7 letters ]+. . . .+€̣ΚΝΟ̣
+. . . .+ΝΝ̣Ο̣Δ̣ . Μ[ 6 letters ]+. . . . . . .+
. +. . . . . . . .+[ . . . . ]+. . . . .+
ΓΥ̣ΝΑ̣Ι[ . . ]+. . .+[ . . . . ]+. . .+ΗΔΟ̣Ν€ΜΟ . Φ
ΔΑ̣Ϲ€ΝΙ+. . . . .+[ . . . . . ]†+.+Ϲ€+.+ . . ΦΦ . Π†
+11 letters+[ . . . . ]+. .+Τ̣^Α€Μ̣Α+. . .+ω̣Ϲ+. . .+Δ̣
+9 letters+[ 6 letters ]+. .+ΠΟΤΤΑ̣ΤΟΠ


Now for the "Why is this nowhere?" part. So on the left we have one single line. Let me calculate which line it goes with of those on the right, and how many letters are in the greens that start the others. 10 letters in the green, last left line matches third never-before-transcribed right line. Let me just transcribe that lonely lefter, so that I can close the second image sooner.

• I guess we start with a 2-letter space;
• Then we have a theta, half-blue half-yellow, average red;
• Then an eta, mostly yellow and yellow-average, with the top of the left leg red;
• Then an alpha, mostly black, with the bottom of the left leg blue;
• Then a mostly-yellow sigma with some small blue and black contributions;
• Then a 1 or 2-letter space, and finally a probable blue alpha – underdot.

That is l. 23, left side. Now for the right side, we start at l. 21.

1. In l. 21:
   • 2-letter green, then a red-average alpha with lots of blue, some yellow, and a little red;
   • Then a blue mu;
   • Then a nu with black left leg, red middle, and yellow right leg, red average;
   • Then what seems like a red subscript alpha corrected into a rotated epsilon which is mostly blue but has some red an yellow and thus averages to red;
   • Then the weirdest blue tau ever – underdot;
   • Then I'd say a yellow omicron tying into a mess which will be a red vestige;
   • Then a mostly-red red-average epsilon with some yellow;
   • Then a red pi, and finally a mostly-yellow iota with a little red at the top.
   And we forsake pic 2 and move to pic 3.
2. In l. 22:
   • 2-letter space, then a black-average alpha with a little red and a black quirk;
   • A clear black nu, a red theta partially overlapping with a mostly-yellow epsilon with a blue bottom;
   • Finally, a half-yellow mu which continues red and has a blue right leg, red average.
3. In l. 23:
   • 3-letter space, then a red-average mostly-yellow pi with a little blue contribution and a black overline;
   • A black omicron, then two red epsilons – actually, a red eta with round legs –, then a 1-letter space, epsilon-omicron in red, and a mu with red left leg and yellow rest, yellow average.
4. In l. 24, the last one, 7-letter space, yellow omicron, blue overlined epsilon, blue sigma, half-blue half-black epsilon with two middle strokes one of each color, I'll say black-average since most of it is black, but underdot for the two strokes.

And here is the transcription:

+10 letters+[ 6 letters ]+. .+ΑΜΝΑ€Τ̣Ο . €ΠΙ
+10 letters+[ 6 letters ]+. .+ΑΝθ€Μ
+. .+θΗΑϹ+. .+Α̣[ 6 letters ]+. . .+ΠΟΗ+.+€ΟΜ
+7 letters+[ . . . . . ]+7 letters+Ο€Ϲ€̣


And now we discuss what this could read as.

1. In l. 1, drop a nu and a iota, add a xi, and you get almost Edmonds' line 2, modulo dropping the end into a locus desperatus; that gamma cannot be a tau since its top doesn't extend to the left of the vertical leg;
2. In l. 2, the beginning, after dropping a mu, is an "oh" plus Edmonds' l. 1's, and the rest is desperatus; φιγε cannot be a misreading of φιλή(σω), because an eta cannot be misread as an epsilon (though a gamma might be a misreading of a lambda if said lambda has the right side very lightly slanted and almsost horizontal);
3. L. 3 starts with πέφαν', dropping the +.+, I'd say, and then ends with ὔδατι δὲ dropping a few letters, and the middle is an undecipherable mess; I think I will be daring and complete ἴσχυν near the end;
4. L. 4 starts with κρίνον and seems to have ἔπε' οἶα, "words such as", near the end, and the rest I don't know; I think I'll see if I can find something with μηρ in it; I believe the second epsilon of ἔπε', which has a red quirk turning it to an alpha, might be the source of περ in LP and παρὰ in Edmonds, but what follows is definitely not a rho, since it has round stuff which in this parchment is not what rhos have;
5. In l. 5, πέπλον is definitely there, and the πυσχ is probably hiding in that +. .+, but, as much as I dislike it (I mean, I always referred to this fragment as "the πέπλον-πυσχ" /'pe.plon 'pysk/), I just cannot see it; anyway, after πέπλον I can read nothing meaningful;
6. L. 6, after dropping a chi, has much of Edmonds' beginning, then I can complete an ἔλοισα and read ὠς at the end;
7. In l. 7, we have the κροκόεντα, and the end, with two small fixes, is also Edmonds, but the middle seems to be λωόρεα, which I don't know if it exists;
8. In l. 8, πέπλον πορφύριον is an obvious completion, and the remaining readable letters spell ἐνδέξω, so nothing special here;
9. In l. 9, honestly I think the phi before the hole could be a blue sigma and a blue-average epsilon, so I update the transcription, and I can now complete to Edmonds' text, adding a tau in the ΝωΝ, where I think there is actually room;
10. L. 10 can easily be completed to Edmonds' text, which has one letter less than marked in the hole, but I just counted those letters twice and ended up with 11 once and 12 the second time, so the estimates are probably rough; the theta above the hole is probably a quirk;
11. OOPS! It seems I transcribed the end of l. 11 both in l. 11 and in l. 12, and I skipped the end of l. 10 in transcribing; that's a big fix to make; the end of l. 10, after the rho, is [. . . . .]+. . .+θ+.+Ν€[.]θ€ΟΝ̣ΗΝ, that is:
    • Hole and space, then the famous black theta, a space, a red nu;
    • A half-blue half-yellow red-average epsilon;
    • A space, then a half-red half-yellow mostly-yellow yellow-average theta, then a red epsilon;
    • The rest is straightforward; I just note the nu has lost its middle so it could be a mu hence the underdot, and the eta has a red quirk on it;
    With that, Edmonds' text doesn't exist, and l. 10 actually ends θέον ἦν, with Gibberish before it; as for l. 11, IA is now bogus, and we get that mess of l. 10, which I'm going to turn to ΟΝ€ . , and leave as is;
12. Now l. 12 has to end in one of two different things; actually, I'll keep the kappa as was, and reinstate that alpha, which is red-average and certain, so we end in Κ̣Α now; well, φρῦσσον ὦ is there, but with the ending not Edmonds' I find it too precarious to complete Πραξίνω;
13. In l. 13, I drop the first greened-out area and read a rho in the following vestige, the next vestige as a theta, and the correction to the green as an epsilon with some red quirkiness closing the "bottom half"; Edmonds' completion has too few letters: I recounted and got 14 for the hole, he offers 8 plus the omicron he has outside the hole but I'd have inside it; so I will read ποήσω at the end, but not complete πότον;
14. In l. 14, I can only complete τέκνον at the end, and maybe read τὺ at the beginning, though that form doesn't convince me since I only recall reading it in Edmonds; the rest is mystery;
15. In l. 15, I can't complete anything;
16. Same goes for l. 16, where Edmonds' text starting with a mu is, I wager, the least likely one;
17. L. 17 definitely starts with γυναίκων, but past that Edmonds abandons me and I can only conjecture ἠδον' ἔμοι and drop the phi, but ἡδονή would probably be ἀδονά in Aeolic, so scratch that; leave it as is;
18. In l. 18, looking again, I see a potential blue epsilon before the delta, but is ἐδαεν going to give me anything? Thanks Perseus! ἔδαε(ν) means (s)he learnt, so here we go; the rest is Gibberish;
19. In l. 19, I guess τὰ ἐμὰ and ὠς can be read, though that hiatus is hardly convincing…;
20. In l. 20, I can read ποττὰ, apparently «adj sg fem nom/voc epic doric aeolic» from πόσος (can also be dual feminine or plural neuter in a "casus rectus" – and I was giggling at it without the stress because "potta" is a dialectal term for vulva :) ), and then the τοπ? Mystery…;
21. In l. 21, I definitely read δάμνατο, so screw that correction, and then ἐπὶ, I think I'll keep the vestige as a chance of avoiding the hiatus;
22. In l. 22, ἀνθεμ is a form of ἄνθεμος chopped off, so I'll have empty angled brackets for it;
23. In l. 23, πάθηα suggests itself, or rather βάθηα, which I'm pretty sure I've seen elsewhere; beyond the hole, ποη suggests a form of ποέω, maybe ποήσομαι, dropping the green, fixing the epsilon, and completing; however, I cannot ignore that middle stroke, so that hypothesis is out the window; I could, however, reanalyse that mu as a nu with quirk, and read ποητέον there;
24. In l. 24, I can do nothing.

So here is the revised transcription with the reordered text to its right.

ωΦΑ̣Ι̣Ν̣Ν̣Α̣Μ̣Μ̣Ι̣Κ[.]Η€̣ . ΝΑΝ̤Λ̤ΥΙΙ€€Γ€ΟΝΑ . ΨΑ̤ΠΦ̣Η̣Μ[]ΜΑ . . Ο+.+ . Π . ΦΙΓ€ . . . Π€ΦΑ̣+.+ΝΙΝ̣[. . .]Ρ̣Ν . . +.+Χ+.+ΝÝΔΑ̣ . Υ̤Δ̄€̣ΔΙΔ€ ΚΡ̣ΙΝΟ̤Ν . . [. . . .] . Μ̣H́Ρ̣ . . €̤Π€ΟΙΑ . . ΑΙ Π€ΠΛ̣ΟΝ̣ . ΙΦ̣+. .+ΧΝ . Ι̣Ι̣ . Γ . Ο[. .]IA ΚΑ̣ΙΚ̣Λ€̣IΙϹϹΑω̣ΧΝ+.+ΛΟ . ΙϹΑ+. .+ω̣Ϲ ΚΡΟ̣ΚΟ€ΝΤ̣AΛω̣[]Ο̣Ρ€.Ϲ€+.+ΒΑΛΤ̤ΗΚΟΙ Π€Π̣ΛΟΝ̣ΠΟΡ̣ΦΥ̣[.]ΙΟΝ+.+ . . €ΝΔ€Ξ̣ω̣+. . . . . . . .+ Χ̤ΛΑΙΝΑΙΠ€ΡϹ̣[. .]ΡΙϹΑ̣Ν̣ωΝ . Ν+. . . . . . .+ ϹΤ€€Φ̤ΑΝ̣[.]ΙΠ€̣Ρ̣[. . . . .]+. . .+θ+.+Ν€[.]θ€ΟΝ̣ΗΝ {ω̣}ΚΑ̣Λ . . ϹϹ̣Α̣ΜΑ+. . .+[. . . . . . . . . . . .]ΟΝ€ . ΦΡ̣ΥϹϹ . ΝωΠΡ[ 12 letters]Κ̣Α ΠΑ̣+.+ . . +.+€̣Νω̣ΝΠΟ[ 18-19 letters ]Η̣Ϲ . . ΤΥΝΑ̣ΚΚ+. .+ΙΠ+.+[ 6-7 letters ]+. . . .+€̣ΚΝΟ̣ +. . . .+ΝΝ̣Ο̣Δ̣ . Μ[ 6 letters ]+. . . . . . .+ . +. . . . . . . .+[ . . . . ]+. . . . .+ ΓΥ̣ΝΑ̣Ι[ . . ]+. . .+[ . . . . ]+. . .+ΗΔΟ̣Ν€ΜΟ . Φ €ΔΑ̣Ϲ€ΝΙ+. . . . .+[ . . . . . ]†+.+Ϲ€+.+ . . ΦΦ . Π† +11 letters+[ . . . . ]+. .+Τ̣Α€Μ̣Α+. . .+ω̣Ϲ+. . .+Δ̣ +9 letters+[ 6 letters ]+. .+ΠΟΤΤΑ̣ΤΟΠ +10 letters+[ 6 letters ]+. .+ΑΜΝΑ€Τ̣Ο . €ΠΙ +10 letters+[ 6 letters ]+. .+ΑΝθ€Μ +. .+θΗΑϹ+. .+Α̣[ 6 letters ]+. . .+ΠΟΗ+.+€ΟΜ +7 letters+[ . . . . . ]+7 letters+Ο€Ϲ€̣

ὦ φα̣ί̣ν̣{ν̣}' ἄ̣μ̣μ̣ι̣ κ{[.]}ἠ‹ξ› ε̣ὔναν̤ λ̤υῖε †εγεονα† {.} Ψά̤πφ̣' ἦ̣ {μ}μὰν †. ο+.+ . π . φιγε . . .† πέφα̣{+.+}ν' †ἰν̣[. . .]ρ̣ν .† ἴ+σ+χ+υ+ν ὔδα̣τ{υ̤δε̣δ}ι δὲ κρ̣ίνο̤ν †. . [. . . .] . μ̣ήρ̣ . .† ἔ̤πε' οἶα †. . αι† πέπλ̣ον̣ †. ιφ̣+. .+χν . ι̣ι̣ . γ . ο[. .]ια κα̣ὶ Κ̣λε̣ῖις σάω̣‹ν› {χν} +ἔ+λο{.}ισα +. .+ ὠ̣ς κρο̣κόεντ̣α †λω̣[]ορε .† σ' ἐ+β+βάλ{τ̤}η κ‹α›ὶ πέπ̣λον̣ πορ̣φύ̣[ρ]ιον †+.+ . .† ἐνδέξω +. . . . . . . .+ χ̤λαίνᾳ πέρ σ̣' [ἐξα]ρισάν‹τ›ων . ν+. . . . . . .+ στέφ̤αν̣[ο]ι πε̣ρ̣[ὶ . . . .]†+. . .+θ+.+νε[.]† θέον̣ ἦν κἄ̣λθ' ὄσσ̣α̣ †μα+. . .+[ 12 letters ]ονε .† φρ̣ῦσσον ὦ †πρ[ 12 letters ]κ̣α† πα̣{+.+}ρθ+έ̣+νω̣ν πο[ 12 letters πο]ή̣σω {.} †τυνα̣κκ+. .+ιπ+.+[ 6-7 letters ]+. . .† τ+έ̣κνο̣‹ν› †+. . . .+νν̣ο̣δ̣ . μ[]+. . . . . . .+ . +8 letters+[. . . .]+. . . . .+† γυ̣να̣ί[κω]+ν †. .+[. . . .]+. . .+η† δό̣ν' ἔμοι {φ} ἔδα̣εν †ἰ+. . . . .+]. . . . .]+.+σε+.+ . . φφ . π +11 letters+[. . . .]+. .+† τ̣ὰ ἔμ̣α +. . .+ ὠ̣ς †+. . .+ δ̣ +9 letters+[ 6 letters ]+. .+† ποττὰϷ †τοπ† +10 letters+[ 6 letters ]+. † δ+άμνατ̣ο . ἐπὶ †+10 letters+[ 6 letters]+. .+† ἀνθεμ‹› +βά+θηά †σ+. . +α̣[]+. . .+† ποη+τ+έον †+7 letters+[. . . . .]+7 letters+οεσε̣†

And with this, I'm finally done with this thing, and the post is complete – except for the next papyrus, which was transcribed in the middle of this hellhole-parchment.

P. Köln inv. 0008r, big single fragment, not sure if Sappho but looks familiar

]+ . . +Π̣ΥΚΔΑ[ ]+ . +ΠΟΔΑΝΑ[ ]. €ĨωΛΟϹΑΤ̣[ ]Χ̣€̣ΙΧΑΡΙ€ϹϹΑΙ[ ] . €ΧΟΙϹΑΘΥΜΟ[ ]. ΑΛΑΜωΙ[ ]€̣ΙΚΡΟΤΑΛωΝΨ[ ]ΤΟΔΥΤωΝΥΜΟ̣[

]+ . . +π̣υκδα[ ε]+ὔ+ποδα να[ ]. ειωλοσατ[ ]χ̣ε̣ι χαριέσσᾳ [ ] . ἔχοισα θῦμο[ν ] θαλάμῳ [ ]ε̣ι κροτάλων ψ[ ]το δύτων ὔμο[ι

… p]retty-feeted [ … ] to the graceful [ ]having hear[t ] for the wedding-chamber [ ] of clappers [ ] of the sunk ones, the same wa[y

In l. 4, the first chi is half-cut, could be a lambda, less likely because a lambda is likely to be taller; the following diphthong has an epsilon which is strangely faded. The final omicron could be a half-omega.

P.GC. inv. 105 fr. 4 and P.Sapph. Obbink: to join or not to join?

Only very few days ago, I found this paper which mentions someone suggested these two papyrus fragments be jioned. Physically speaking, there isn't much evidence to the joint, given that only one line would join, and the margins are pretty straight. Let's see the combination, with P.Sapph. Obbink uncolored, P.GC. inv. 105 fr. 4 blue, and P.Oxy. 1231 fr. 16 red:

Πῶς̣ κε δή τις οὐ θαμέω̣ς ἄσαιτο,
Κύπρι, δἐσπο̣ι̣ν̣', ὄττινα̣ +δ+ὴ φίλ̣[ησι
] θέλοι μάλιστα πάθαν χ̣άλ̣[ασσαι;
]ονεχησθα
πᾷ [σ]άλοισί μ' ἀλεμάτ̣ως δαΐσδ[ης
εἰ̣μ̣έρ‹ῳ› λύ{ι̣}σαντ̣ι̣ γόν' ὠμ . . . [
πολ[.]απα . [ . . ]αμμοὐ προ̣[
οὖ̣τον ἔ̣ερθαι


So the iota in πᾷ is what became of the vestige in GC fr. 4, and then a lacuna is left. The supplements are somewhat convincing, though one needs that unconvincing reading to work. In ll. 6-7, people agree there's more to be read. In the first case, I can see room for 2-3 more lettertops, so I added three dots, but they read 4 and maybe an extra one. In the other line, after προ, the papyrus ends. So either there's a fragment joined to the right I'm missing, or GH missed something in the Oxyrhynchus fragment, or WTF are those letters from. Many of the readings in the critical apparatus of the paper have γόνωμ', but WTF is this word? Seems it should mean "I invoke", but nothing I've tried looking up has resulted in anything. Oh, Aeolic for γουνοῦμαι, which actually comes from γόνυ and means "hug the knees of, beg"? That sounds convincing. I think l. 3 stays as was, l. 4 becomes πῶς] ὀνέχησθα, l. 5 only changes its start, l. 6 ends following the article, and let's see ll. 7-8. First, though, this λύσσαντι idea in l. 6 could work, but reading a sigma there is pretty forced IMO, I'd bet for an iota, and indeed I made it certain in my transcription. So that stays. We can still keep that high dot after it, if we imply an object, e.g. "me". πόλλα starting l. 7 seems unescapable. πάμπαν can then be read, if we assume the first readable mu is assimilated (cfr. νᾷΐ σὺμ πλήᾳ at the beginning of the papyrus), so we have πάμ[π]α‹ν›. The ερης others read is unfindable to me, but I can still supplement it. I'll take the two-letter blank idea and complete the combination as follows:

Πῶς̣ κε δή τις οὐ θαμέω̣ς ἄσαιτο,
Κύπρι, δἐσπο̣ι̣ν̣', ὄττινα̣ +δ+ὴ φίλ̣[ησι
κωὐ] θέλοι μάλιστα πάθαν χ̣άλ̣[ασσαι;
πῶς] ὀνέχησθα

πᾷ [σ]άλοισί μ' ἀλεμάτ̣ως δαΐσδ[ης]
εἰ̣μ̣έρ‹ῳ› λύ{ι̣}σαντ̣ι̣; γόνωμ' ἄνα[σσα,]
πόλ[λ]α· πάμ[π]α‹ν› μ' οὐ προ̣[τέρᾖσθ' ἀπέχθης,
οὔ̣τ' ὀνέ̣ερχ+θ+αι

Note that the blank would also include the alpha of the -αι ending, but that ending is preserved in the Oxyrhynchus fragment. In other words, ουτο + νεερχ+θα+ι + αι = οὔ̣τ' ὀν ἔ̣ερχ+θ+αι. The problem here is that infinitive at the end, governed by nothing. Obbink 2014 (quoted by the paper) says «The traces visible above the line here are not parts of a suprascript iota after ] ̣ερ, but rather the remains of offsets left when the scrap bearing ] ̣ερησ[, originally folded over left, was unfolded and turned back into place during conservation». I don't get it if it's referring to the P.Sapph.Obbink, which seems to clearly have ink tracing a half-iota, and if it's about P.Oxy. 1231 fr. 16, well, I have no image. GH give αι with no doubts and no notes. The P.Sapph. Obbink trace is also curved rightward, so it could never be a sigma, as the paper would have it. So everything points to the paper's ὀνέερξας being contra papyrum. ὀνέερξαι could also be read, the xi being the vestige (small dot is its top, the rest, which is most of it, lost to the blank), and the alpha again in the blank and found in P.Oxy., so ὀνέ̣ερξαι. Still, what governs this infinitive? Well, actually, ὀνέερξαι could also be an imperative, and we could add Νῦν με in the next stanza to complete the sentence. That's it.