Tracing mysterious line ends in a combination of Sappho fragments found in an Italian anthology

Question

Somewhat recently, I stumbled upon this Italian Sappho anthology, where, among other combinations, the following is found:

First off, they are not listing all the fragments joined into this combination. Besides LP 86, LP 60, and LP 65, there are also LP 66(c) and LP 67(a) in here. I tried to trace all I saw here, and mostly succeeded, except for ll. 12-13, which seem to have come from nowhere. I color-coded LP 60 black, LP 67(a) blue, LP 65 red, LP 66(c) green, the unidentified end-of-lines pink, and LP 86 purple, and, with minimal adjustments to the text (including reinstating a completely ignored line of 67(a) which gives the very beginning of l. 21 and fixing the inmetrical ἀμφιβιβάσκει in l. 27 to ἀμφιβάσκει, hopefully without altering the meaning too much - and the translation "circonda", "surrounds", makes me optimist since I know ἀμφιβάσκει means "surrounds" from an Edmonds restoration reading μυρίαν ἄμμω]ν ῤύτιν ἀμφιβάσκει), I got the following:

Note that the supplements are usually left black, unless they are somehow "called for" by a papyrus fragment, in which case they take the color of that fragment. Note also that this is a preview of a blog post due next Saturday, hence the green background. Finally, note that I introduced the notation of underlined letters for those letters that, in a raw transcription, would be simple "."s, but are guessed based on surrounding words/letters or on accent marks above them.

I have a couple guesses as to the origin of those pink letters:

1. They come from lines in P.Halle 3 (source of the black part of my collage) which for some reason weren't included in the Lobel-Page text. Which would clearly beg the question: why did LP not include those lines? And why did this anthology do it? Then again, P.Halle 3 was republished as P.Oxy. 1787 fr. 44, and the volume of "The Oxyrhynchus Papyri" containing P.Oxy. 1787 (vol. XV) doesn't have those lines, so I guess we can exclude this.
2. They come from another LP fragment, in which case they'd somehow have to come from LP 87(15), the only one from P.Oxy. 1787 vaguely matching those things. Which would beg the question: where is l. 1 of that fragment, and where are the vestige in «.ν» and the uncertain alpha in the following line come from, since LP didn't have them in the text of 87(15)?
3. They come from another P.Oxy. 1787 fragment which was relocated among another author's works, e.g. one of the hundred-odd that Lobel decided were Pindar (cfr. this older question of mine). Which begs the questions: (a) Was it just two lines? (b) If so, why was it relocated? (c) If not, where is the rest of it in this joining? Looking at The Oxyrhynchus Papyri vol. XV, I realized 87(15) was not P.Oxy. 1787, but P.Oxy 2166(d), so those may have to be searched too. The Oxyrhynchus volume doesn't give any fragments that could be the source of those pink lines, and LP has all 2166(d) fragments from 1 to 13, so either it's 2166(d) fr. 5 (aka LP 87(15)), or it's a 2166(d) fragment with number not between 1 and 13, or I'm missing something.
4. They come from a very recent fragment I have never heard about, found after Lobel-Page. Which obviously makes me curious to know more about it.

Can you help me trace those to their source?

UPDATE

Una Mitra per Cleis, where Ferrari published the fusion's text (or almost - cfr. this other question and my answer there), has those line endings just pop up out of nowhere, as if they were from one of the fragments joined into this. So the mystery thickens.

Answer 1

That same Ferrari book, had I checked further, would have yielded the following note:

In comparison with the reconstruction given in Ferrari 2005b, the text as given here takes into account two contributions of Puglia: a) the combination of F 60 (= P.Halle 3) with fr. 3, hence the necessity of postulating a five-line lacuna between F 86 and F 60 and of considering F 86.1, not F 86.2, as the initial line of the poem; b) the location of F 67 at the left and on line with F 60.1. On this see Puglia 2007: 22–28.

Ferrari 2005b is this article, where the blue and pink parts in the collaged text above are missing, and the lacuna is not postulated.

Puglia 2007 is «Puglia, E. “Per la ricomposizione del quarto libro dei canti di Saffo (POxy. 1787).” SemRom 10 (2007): 17–39», which I can't seem to find online.

The question is: what is "fr. 3"? P.Oxy. 1787 fr. 3? But that wasn't joined into here:

• LP 60, the black part, is from P. Halle 3;
• LP 65, the red part, is from P.Oxy. 1787 fr. 4;
• LP 86, the purple part, is from P.Oxy. 2166(d) fr. 1;
• LP 66(c), the green part, was published as fr. 7A(c) in Lobel's Σαπφοῦς μέλη book 4;
• LP 67(a), the blue part, comes from P.Oxy. 1787 fr. 5.

P.Oxy. 1787 fr. 3 is the source of LP 61-63, none of which was joined into here.

Wait though. P.Oxy. 1787 fr. 3 has a scanty first column, the poems I mentioned coming from col. ii. And on col. i, there is a line-final Ν halfway through the papyrus. Below it, a faint trace of quite possibly an alpha. So I think we have our source.

I tried juxtaposing P.Oxy. 1787 fr. 3 and the purple part, both of which I have an image of, and it seems there is a kinda snug fit between the two. If that is correct, the last line of the purple joins with the first line in the following image:

So the pink line ending in nu would match with τέαυτ in fr. 3 col. ii. But I see no nu to the left of that. Yeah well, dumb me, the purple lines are middle, they can't join with blank ends.

Also, here's the original transcription for fr. 3:

The nu I thought I saw before is where they read ". I". The "." is potentially the top of a vertical leg, which would mean II or N or M. It could be argued that this is ". N", but the right half of such a nu would be so faint I can hardly support it.

It would actually look like the ΗϹ line in col. i should join with l. 6 of P.Halle, where the book by Ferrari ends with θελήσηις while the anthology's text didn't get the memo and still ends in θελήση[ς. This would mean the vestige in the next line is treated as an iota, which can be supported, and the two vestiges in the line after that are dismissed. This is kind of not really supportable IMO. Treating the vestige before the HC line as a high dot, perfect. Problem: this makes P.Halle join perfectly with fr. 3, leaving no lines above it. So… where are these lines from again?

Hmm, I guess this isn't really an answer… this origin isn't convincing at all. Basically, the only convincing joint eliminates any trace of those pink lines because, while fr. 3 extends 2.5 lines higher than its first transcribed line, those lines don't show any trace of ink to my eyes. Plus, the length of the lacuna is completely unjustified.

UPDATE

The answer is presented in this other answer of mine.