I had a look (long one, but not necessarily good) at P123 as published recently in the Oxyrhynchus volumes (we mentioned this one and P124 here).
Text from 1 Cor 14 and 15, IVth century.
The pictures can be found here.
The transcription by Thomas struck me as somewhat optimistic:
μα]ν̣θανωσιν κ[αι παντες
παρ]α̣κ̣α̣λ̣ω̣ντ̣[αι και] π̣να προ̣φητων̣ [προφηταις
υ]π̣ο̣τασσεται̣ ο̣υ γαρ εστι[ν α]κ̣ατασ[ιας ο θς
αλλ ειρ]η̣νες οω εν πασαις τα[ις] εκκ̣[λησιας
των αγι]ω̣ν̣ α̣ι̣ γ̣υναικες εν [ταις εκκλησιαις
σιγατωσαν ου γα]ρ̣ επιτρεπ̣[εται αυταις λαλειν
υ]μ̣ιν εν π̣ρωτοι̣[ς ο και παρελαβον οτι
χς απεθα]ν̣εν υπερ των α̣[μαρτιων ημων κατα
τας γραφ]α̣ς και ο̣[τ]ι̣ ε̣τ̣α̣φη και̣ [οτι] εγηγε̣ρτα̣[ι τη
ημερα τη τριτ]η̣ κα̣[τα] τ̣ας̣ γραφας κ̣αι οτι ωφ[θη κηφα
ειτα τοις δωδεκα επειτα] ω̣φθη επανω̣ πε̣ν̣[τ]α̣
κοσιοις αδελφοις εφαπαξ ε]ξ̣ ων οι πλ̣[ειονες
μενουσιν εως αρτι τινες δε εκο]ι̣μ[ηθησαν
I would be much more cautious and would transcribe the text as follows:
[c. 4]ν.[ ... ]ν̣α̣προ̣φητω.[...
] ... [... ]υ̣να̣ικεσεν̣[...
] . . [
The main difference is that I would only print a letter as secure if it is almost completely visible and would dot those letters that are incomplete but still likely. Consequently, I have quite a few letters less than Thomas in his transcription. Yet, I have no issue with any of the letters or, by and large, with the reconstructed text (perhaps ιβ rather than δωδεκα).
The only real question I have if the two strokes above the ν are a scribal marking (recto, line 1), (but what?):
It is quite handsome that one can use NA27 as the base text for this papyrus. The only possible textual variation then is whether the nomen sacrum πνα can be seen as the plural πνευματα (Thomas refers to Paap for support). Difficult to decide without knowing the scribal patterns in this particular manuscript.
By the way, I think the fragments as they are on the Oxyrhynchus website are slightly displaced. Not much, but enough to be wrong-footed trying to squeeze all the letters into the blank spaces.
My version would be more like this: