In November 2014 I had the pleasure of examining Bruxellensis 11377–80. This paper manuscript was owned by Pierre Pantin at his death in 1611; it occupies no. 30 in his catalogue, where the first text, on fols 1–4, was misidentified as ‘Hymni Homeri’. The next owner, André Schott, repeated the error; Omont (and following him the Pinakes database) uses a page-numeration that completely excludes the current fols 1–4.Footnote ¹

The manuscript is composed of three parts, as follows:

1. 1) fols 1–4 in fact contain Proclus' Hymns 3.12–7.52, together with Orphic fr. 31 Bernabé = 21a Kern; there are twenty-two lines per page. This source is absent from Vogt's edition, even though fols 92v–95r, which contain a further text of Proclus' Hymns, are included as MS 5.Footnote ² I identified the scribe as Demetrios Moschos by his handwriting.Footnote ³ The watermark is a crossbow within a circle; the chain-lines pass just outside the circle, c.32 mm apart. The design is similar to Harlfinger's ‘Arbalète 35’, but smaller. The watermark described by Formentin from Marcianus gr. 604 fols 132–78, where Moschos copied the Orphic Argonautica, may offer a better match.Footnote ⁴ I do not know why fol. 1 is also marked ‘29’ in a later hand.

2. 2) fols 5–30 contain the Theognidea, with twenty-four lines per page. The main watermark is a close match for Harlfinger ‘Main 29’ (dated c.1490).Footnote ⁵ A further watermark (fols 7/10) will be described in the next paragraph. The scribe is, I believe, the same as that of Parisinus gr. 2765 fols 39–58 (Homeric Hymns), which is a twin of the Bruxellensis' text of the Homeric Hymns. The range of scribes collaborating on the Parisinus, including Georgios Gregoropoulos and Thomas Bitzimanos, points to Apostoles' workshop, which is the source of the main part of the Bruxellensis (see below).Footnote ⁶

3. 3) fols 31–96 contain the Homeric Hymns, Moschus' Amor Fugitivus, the Orphic Hymns and five Proclan Hymns (1, 2, 3–4 run together, 5), with twenty-five lines per page. The numbering of the gatherings reverts to 1 here. This section is signed (fol. 95r) and mostly written by Aristoboulos Apostoles.Footnote ⁷ He signs himself as a deacon, a position that he already held in March 1489.Footnote ⁸ The watermark is a close match for Harlfinger ‘Main 11’ or ‘Main 17’, and related watermarks from the early 1490s.Footnote ⁹ The collocation with ‘Main 29’ (above) suggests a connection to Vaticanus gr. 1311, which Harlfinger cites for both his ‘Main 11’ and ‘Main 29’, and which was written by Aristoboulos c.1490.

Collating fol. 3r allowed me to locate Moschos's text of the Proclan Hymns in the stemma. The range and order of hymns put the manuscript within Vogt's families ε or δ, and the reading 4.1 σοφίης reduces this to ε or β (since γ, the other sub-family of δ, has σοφίαν).Footnote ¹⁰ However, sporadic readings appear to be incorporated from the O-family, which contains Plethon's idiosyncratic ‘edition’ of the hymns. Particularly revealing is 7.51-2:

Here βε have:

For O, Plethon conflated the lines into one: κέκλυθι κέκλυθι καί μοι μείλιχον οὖας ὑπόσχες. If Moschos was conscientiously collating two manuscripts to produce his copy, he must have miscopied O for verse 51 and βε for verse 52, to produce two lines with identical second halves, rather than accepting the superior text of βε. A more likely scenario is suggested by Mutinensis Estensis α.W.5.16 (= gr. 164). This manuscript is a member of β, which Giorgio Valla copied initially in 1464/1465, before subsequently collating his text of the Hymns with a copy from the O-family.Footnote ¹¹ Vogt assigns this manuscript siglum B, its marginalia B². Valla wrote in the margin to 7.51 ἢ καὶ μοι [sic] μείλιχον οὖας ὑπόσχες, and I believe that Moschos was misled by trying to incorporate precisely this marginal note.Footnote ¹² Orphic fr. 31 also follows in B; in the margin to line 1 appears the unusual variant ἀρχικέραυνος, which Moschos incorporates. One might argue that Moschos's omission of Proclus, H. 6.13-15 does demonstrate direct access to an O-manuscript as well as to B², since (based on Vogt's information) B² appears not to mark those lines for deletion.Footnote ¹³ If this is reliable, Valla at around the time of B² may have been Moschos's source for both manuscripts.

That Moschos should have copied Proclus' Hymns, from a book owned by Valla, is no great surprise. He copied other works of Proclus in Mutinensis γ.V.4.4 and Parisinus gr. 1840; Proclus' Hymns are transmitted alongside Orphica, and Moschos copied the Orphic Argonautica in Marcianus gr. 604 and Vaticanus gr. 37, as well as writing introductory notes to the Lithica for Pico della Mirandola. He wrote Mutinensis α.P.5.18, which Valla owned, and he probably also knew the heir to Valla's library, Alberto Pio of Carpi.Footnote ¹⁴ Moschos's life is too obscure for us to specify the circumstances of copying, but if I may hazard a speculation it would be based on two facts about Pico. The first is that Pico wrote to Valla in May 1492 asking to borrow B (though his stated interest there was Callimachus' Hymns).Footnote ¹⁵ The second is that Pico, according to catalogues of his library made after it passed to Domenico Grimani, owned a book containing three of Proclus' Hymns—a unique quantity, as far as I know.Footnote ¹⁶ It may be that the catalogues' ‘three’ was a mistake. However, Pico's book, in which the range of texts is not matched elsewhere, presumably had multiple sources. Hence it is conceivable that Moschos copied the Orphic Argonautica and Hymns and Proclus' Hymns for Pico, using Valla's copy; the last couple of pages went astray, so that this text contained only Hymns 1.1–3.11, which were later bound with the Homeric Hymns, Musaeus and Aratus from other sources. The errant pages of Proclus somehow ended up in the Bruxellensis, alongside texts from Apostoles' Cretan workshop, with which Moschos is known to have had dealings via Venice in the 1490s.Footnote ¹⁷ My speculation is, admittedly, only one possibility; since Pico's book was probably destroyed in a fire in 1687, it is hard to see how it could be confirmed. And so I leave matters there.