Marcus Musurus’ annotations in the ‘recovered’ UCM 22 (= 442 rahlfs) manuscript: Maccabaei II

: The annotations of the Greek humanist Marcus Musurus in the recently recovered biblical manuscript UCM 22 (= 442 Rahlfs, 16th century in .), corresponding to the Septuaginta ( Maccabaei , Book II), are studied in this article, as well as their possible origin: the Marc.gr. 1.

T he Universidad Complutense de Madrid (UCM) BH ms.22 Villa-Amil (= 442 Rahlfs) (abbreviated, UCM 22), from the Historical Library "Marqués de Valdecilla" at the Complutense University, is a Greek codex on parchment, which was apparently copied in the late 15th or early 16th century.1It contains part of the Old Testament (Septuaginta) and for a long time was considered "deperditus"2 in the fighting during the Spanish Civil War (1936)(1937)(1938)(1939) in the University City of Madrid.The manuscript was likely sent by the "Signoria" of Venice to Cardinal Cisneros in the early 16th century to serve as a model for the Greek column of the Biblia Polyglota Complutensis, which the cardinal was promoting in Alcalá de Henares and which was the first printed edition of the Bible, chronologically speaking, although it was distributed throughout Europe shortly after the Aldine edition.3Originally the manuscript, on parchment, 370 x 250 mm., contained 307 ff and ended with Maccabei III.4Some 600 fragments are currently preserved, to a greater or lesser degree of conservation.
The recent ordination and digitisation of all the preserved fragments of the UCM 22 manuscript,5 which we have been able to carry out thanks to the incorporation into the project of our Portuguese colleague Carlos Martins de Jesus,6 has allowed us to verify that, as already suspected,7 this manuscript was a model for the Polyglota, especially in vol.IV (Maccabaei) and, to a lesser extent, in vols.II and III.
The "recovery" and subsequent digitisation of all the preserved fragments of UCM 22 has also revealed the many annotations made by the well-known Greek humanist and collaborator of Aldus Manutius, Marcus Musurus (1470-1517),8 who also annotated, although more scarcely, the Lond.BL Add.10968 manuscript.This is considered to have been the model for the Aldine9 edition and with it the Complutensian manuscript seems to have a close textual relationship, perhaps because it shares the same model: Marc.gr.5,10 itself a copy of Marc.gr. 16. 5 Accessible at: http://dioscorides.ucm.es/proyecto_digitalizacion/index.php?doc=5309456614& y=2011&p=1 See also: https://www.youtube.com/watch?v=QeY5VLChq4c.Before its partial destruction during the Spanish Civil War (1936War ( -1939)), only three passages from the manuscript were known.Two of them (Iudices,(19)(20)(21)(22)(23)(24)(25)(26) were copied at the end of the 19th century by the scholar Pascual Gayangos and were reproduced by Delitzsch (1886, 13-17 and 17-18).The other passage (Reges II, 23.1-5) was transcribed by Revilla (1917, 100-101).The current digitisation, which is now almost complete, would not have been possible without the contribution of the colleagues who preceded us with their valuable work on the manuscript, such as Antonio Bravo García, Natalio Fernández Marcos and Jesús Ángel Espinós.To all of them, we pay a tribute of sincere gratitude.
6 In the research projects "Greek manuscripts in Spain and their European context" (II) FFI2015-67475-C2-2-P and (III) PID2019-05733GB-I00, with the technical support of the staff of the Historical Library at the Complutense University.We would like to pay tribute to all involved, with a very special remembrance going out to the humble caretaker of the Faculty of Philosophy and Letters of the Complutense University, L. Ángel López Castro who, during the turbulent years of the Spanish Civil War (1936War ( -1939)), was apparently responsible for the "salvage" of some of the most precious bibliographic treasures that are now housed in the Historical Library at the Complutense University (De Ontañón, 2005).
9 Cataldi Palau (1998, 451, 458) recognised printing marks on the manuscript by the Aldine editors.These two manuscripts, Lond.10968 and UCM 22, would be the only biblical manuscripts annotated by Musurus.The Lond. 10968, whose copyists have not yet been identified with certainty, although the names of Bartolomeo Zanetti, Constantinos Mesobotes and Demetrius Ducas have been mentioned (Cataldi 1998, 459;Speranzi 2013, 271;Jesus 2020a, 725, n. 34), was probably also copied in Venice at the beginning of the 16th century.The manuscript was "a" model of the Aldine edition, but was not, pace Cataldi, "the" only model.In fact, in recent works (Hernández Muñoz 2020a, 231, n.6;Jesus 2020a, 741-742) we have suggested the possibility that the two biblical manuscripts that were annotated by Musurus, being to a certain extent complementary in their content, served, in principle, for the creation of the Aldine edition, in which Musurus collaborated.UCM 22 was then sent to Cardinal Cisneros, promoter of the Polyglota in Alcalá de Henares, as an aid for the work of this edition.In any case, the question of the sources of the Aldine edition of the Greek Bible remains still not fully resolved with there being different models according to the biblical books and even "eclectically" within the same book or passage (Hernández Muñoz 2020b).

The text of UCM 22 and Musurus' annotations
For each passage de Maccabaei II we copy the text transmitted by UCM 22 (ff.189v-202v.)and in brackets and italics ( ) the annotation of Musurus to facilitate its location in the passage.the so-called "Anonymus ΔΤ" (Martínez Manzano & Hernández Muñoz 2019, 142).It is possible that the "castigatissimo codice" to which Cisneros refers in the "Prologue to the reader" of vol.I of the Polyglota refers not so much to Marc. 5 but to its copy, UCM 22, which was revised by Musurus.
11 For the details of its chronology, see Andrist (2020, 46)  19 onwards, we have been able to include the testimony of Lond.10968, which currently does not transmit the previous part.In its present state, Marc.16 also shows some mutilations and transpositions of folia; therefore, its testimony could not be included in all the passages, as in 13.2 or 14.3, where it would probably coincide with a copy of its own: Marc. 5. 15 ex 9.20.Figure 1 can be seen, with two annotations by Musurus in the margin.The images have been reproduced with permission of the Historical Library at the Complutense University.
16 Below ἐν there is a circular stroke indicating a correction of Musurus, which is probably his indication of omission, as it happens on Marc. 1 compared to the rest of the testimonies, including Polyglota.17 after τέσσαρας there is a ^ mark, which indicates that there was an annotation by Musurus in the margin, although not preserved due to damage in the manuscript, cf.12.11; it was probably καὶ εἴκοσι.
18 ut vid., because Musurus' annotation is not very visible today.19 after νομάδες there is the ^ mark, which (see supra, 10.33) indicates an annotation in the margin which has not been preserved today due to damage to the manuscript.It was probably ἄραβες.
20 The damage to the manuscript does not allow us to read all of Musurus' annotations on the lines left blank in UCM 22. Three of the collated manuscripts, Marc.5, UCM 22 and Lond.10968, have the same blank lines.On the other hand, Marc.16 also omitted the text, but this was later added in a barely legible manner in the margin (see also 5.14).This would be evidence that Marc. 16  for Marc. 5 (which he could not read or consider it a marginal gloss, and therefore left the blank space for the added text of Marc. 16), and that Marc. 5 was the model for two "sibling" codices: UCM 22 and Lond.10968, both annotated by Musurus (Hernández Muñoz 2020a, 245, n. 51).However, here UCM 22 is more complete than Lond.10968, a manuscript that does not have the Musurus addition.On the kinship of these manuscripts see also 10.37 (om. τούτου); 11.13 (ἑαυτοὺς instead of ἑαυτὸν); 14.3 (om. τὸ), cf. also 7.40 and 12.34.The UCM 22 text inserted in [ ] is conjectural; the rest is effectively preserved.In Figure 2, lin.12-13, there is the addition of Musurus in the blank space.In lin.15 and lin.22. the s.l. additions can be seen, with the typical ^ mark.
21 παρατα[ξαμένους appears to be written by Musurus in the space left blank by the UCM 22 copyist.22 In Figure 3 Marc. 1, 2, 4, Polyg., om. τὸ Marc. 5, Lond. 10968, Ald. et UCM 22a.c. 32 14.7 [ἀρχιερωσύνην], δεῦρο (s.l. δεύτερον) νῦν ἐλήλυθα: Marc. 1 14.8 δεύτερον δὲ καὶ τῶν ἰδίων (s.l. ἡμετερων) πολιτῶν: Marc. 1 14.11 προσκαλεσάμενος (marg.προχειρισάμενος): Marc. 1 From these data, it can be concluded that the Musurus' annotations in UCM 22 (in Maccabaei Book II)33 seem to come from Marc. 1, a manuscript in capitals from ca. 8th century, since a significant amount of its text only seems to be found, among the manuscripts collated in our study, in this vetustissimus manuscript from the 30 As already noted, παρατα[ξαμένους] seems to have been written by Musurus in a space left blank (or erased) by the copyist of UCM 22.What is offered here by Marc.16, Marc. 5 and Lond.10968 is περιταξαμένους.Of the collated testimonies, only Marc. 1 and the Aldine edition omit πρὸς τοῦτον δὲ, though in the following word there are differences: παραταξαμένων (Ald.), παραταξαμένου (Marc. 1).The continuation of the word in the preserved part of UCM 22 has been lost.Most probably it was παρατα[ξαμένους], as in Polyg., Marc. 2 and Marc. 4, but this is without ruling out παρατα[ξαμένου], as in Marc. 1.The letters παρατα[ ] appear to be written by Musurus (see also the stroke of ρα in ὀχυραί in the Musurus addition in lin.12 of Figure 2), which are distinct from εὐχαριστήσαντ[ ] (in lin.18 of Figure 2) by John Severus.
31 Both UCM 22 and Lond.10968 have the marginal addition, τριακοσίους, which is missing in the probable model of both: Marc. 5.At present, this passage has been not transmitted by the probable model of the latter, Marc.16; we do not know, therefore, whether it would have also omitted τριακοσίους or added it marginally.This is an annotation which, as in 5.14 and 12.27, would not have been incorporated into the text of Marc 5, with the space being left blank.
32 The Aldine edition has been added to this group of closely related codices.33 In book III, transmitted by a smaller number of manuscripts, there are not so many corrections and annotations in UCM 22, but the available data (Hernández Muñoz 2020a, 241-246) also seem to confirm a connection with Marc. 1 (and in (Andrist 2020, 45).This is a hypothesis that is currently being tested by Carlos A. M. de Jesus, to whom I am grateful for the information.The data presented here suggest such access in Maccabaei, at least, to Marc. 1 (and perhaps also Marc. 4) and would thus complete the information provided by D. Speranzi in his monograph on Musurus (2013).It is true that in the critical apparatus other editions of Maccabaei (Book II), such isolated readings of Marc. 1 are also found in a small group of biblical manuscripts, such as: n. 55 (Vat.gr. 1, 10th century), n. 447 (Ambros.267 inf., a. 1568) and n. 771 (Batopaed. 290 + 113,Par. gr. 682,Mosc. 346, also from the 10th century) in 10.33;n. 58 (Vat. gr. 10,11th century),in 14.7,9.21 (with n. 771) and in 14.3 (here also with n. 771 and with n. 55); n. 771 in 11.1, and n. 347 (Athos 29, 13th century) in 9.4 and 9.19.In other words, it seems that these manuscripts (n. 55, 58, 347, 771) would also be connected to Marc.1; however, the simplest hypothesis seems, in our opinion, that Musurus accessed Marc. 1 (or a copy of it) in Venice.35 See the first example (3.15), although it is a trivial addition, and the doubts about 12.34.36 As well attested in 1515 the Diarii de Marino Sanuto (XX, 177 [100]) apud Ross (1976, 42-542, n. 78), In one example (13.2), it can be seen that the same marginal correction by Musurus in UCM 22 is also found in the margin of Lond.10968, the other biblical manuscript annotated by Musurus.The annotations, at least in the case of UCM 22, would have been made before 1516, the year in which Musurus left Venice for Rome.
37 Until now, it was known that the editors of the Polyglota had used two Greek manuscripts from the Vatican Library that Pope Leo X lent to Cisneros: Vat.330 (= 108 Rahlfs, 13th century, especially for vol.II) and Vat.346 (= 248, 13th-14th century, especially for vol.III).Today, we know that they also used two manuscripts which are preserved today in the old collection of the Complutense University of Madrid: UCM 23 (= 1670, 16th century in., also for vol.III, but only in Psalms) and UCM 22 (= 442, especially for Maccabaei in vol.III), the manuscript with which the present work is concerned.We have recently brought attention to the coincidences between the Polyglota and another manuscript in the Vatican Library: Vat.348 (= 671 Rahlfs), 15th century, a manuscript which seems to be a copy of Marc. 4 (Kappler & Hanhart 19594: 9;Hanhart 19602: 12, n.4), although it may also offer some different reading, as in 4.19 (πρoκομομίζοντεs / παρακομομίζοντεs), 10.5 (τοῦ μηνός / τοῦ αὐτοῦ μηνός) and 12.3 (τηλικοῦτον / τηλικοῦτο).Vat.348 today contains, among the Septuaginta books, only Maccabaei I-III, a part missing in the other two Vatican manuscripts, and we have explored the hypothesis that perhaps the manuscript (or a copy of it) was also on loan to Alcalá.Specifically, for the collated part of Maccabaei II, the readings of the Polyglota always coincide with UCM 22 (either the readings ante correctionem or the readings corrected by Musurus) and, when this is not the case, it coincides with Vat.348, as in 10.6 (μεμονώμενοι), 11.36 (προσανενεχθῆναι), 12.3 (τηλικοῦτον) and 14.5 (καιρὸν δὲ).Thus, if we are right, we believe we have shed light on a question about the models of vol.IV of the Polyglota regarding which their greatest connoisseur, S. O'Connell (2006, 146), considered them to be "totally in the dark".Moreover, thanks to the use of UCM 22 and, more hypothetically, Vat.348, it can be said that the Complutensian editors would have had access, albeit indirectly, to some important manuscripts from the "Cardinal Bessarion's collection" in the Biblioteca Marciana in Venice: through UCM 22, at least to Marc. 5 (its model) and Marc. 1 (corrections and additions by Musurus); thanks to Vat. 348 (if our hypothesis is correct) also to Marc. 4.
different annotations can be seen s. l. vel in marg.with the sign γρ.23 Marc. 5 leaves space blank here, as does UCM 22 (before Musurus' annotation), and the text is added to the margin inMarc.16,as in 12.2724 οὗ and -αξαν are offered only by UCM 22, Marc.16 and Marc.5, which is proof of their connection..34πρὸς τοῦτον δὲ (παρατα[ ] i. v. ut vid.)30 δὲ συνέβη πεσεῖν 13.2 πεντακισχιλίους (marg.τριακοσίους)καὶἐλέφαντας:Marc.1,2, 4, Lond.10968mg31,Ald., Polyg., om.τριακοσίους Marc. 5  14.3 τις προγεγενóμενος (s.l.προγεγονὼς)ἀρχιερεύς: Marc. 1 14.3 [οὐδὲ πρ]ὸς (s.l. τὸ) ἅγιον θυσιαστήριον: They were probably also transmitted by Lond.10968, but this part today is not in the manuscript and cannot be verified.Again, we can see a coincidence ofMarc.16,Marc.5andUCM 22 a.c.(and probably also with Lond.10968).12 this book perhaps also withMarc.4).The reading with the correction in UCM 22 appears first; then, after /, the text of the other testimonies:Biblioteca Marciana,34, although Musurus may also have consulted others, such asMarc.2.35As is well-known, access to Cardinal Bessarion's codices in the early 16th century was not easy(Jesus 2020a: 728, n. 43; 2022: 374, n. 20), but not impossible either, at least for Musurus in Venice.36It is also noted that some of Musurus' corrections in UCM 22, which are here only found in Marc. 1, are also found in the Polyglota text, because its editors used UCM 22 with Musurus' correction and incorporated it into their edition37.Finally, Marc.16 is, as was supposed, a model forMarc.5 and this, inturn, is a model for Lond.10968 and UCM 22.Of these, UCM 34 It remains to be seen what happens in other biblical books not transmitted by Marc.1, for example in Reges. is most likely that most of the annotations from UCM 22 are found in Vat.gr.2106, which originally formed a single codex with Marc. 1