No CrossRef data available.
Article contents
N. HEINSIUS'S FRAGMENTVM CAESENAS OF OVID'S METAMORPHOSES REDISCOVERED1
Published online by Cambridge University Press: 20 April 2016
Extract
Among the manuscripts of Ovid's Metamorphoses used by N. Heinsius (1620–1681) and as yet unidentified or given up for lost is the so-called fragmentum Caesenas (Cs), the collation of which was not carried out by Heinsius himself but provided for him by the Hamburg jurist Lucas Langermann (1625–1686), who was a correspondent of Heinsius, Gronovius and Vossius, among others. According to M.D. Reeve, he was also responsible for adding these notes, using the siglum c, to Oxon. Bodl. Auct. S.V.5, which also includes the collations—by another hand—of A (= Vrbinas ueterrimus, our V2) and B (= Berneggerianus, our P2). The variants provided by this fragment affect lines 9.235–11.169 and 13.1–403, although this does not imply that these were the exact limits of its content, as we shall see below. Some of these variants ended up in the notes of the editions by Heinsius himself and by those who continued his work (the most prominent example being the admirable edition of Ovid by P. Burman, in 1727), and it was D.A. Slater who rescued the collation of Bodl. Auct. S.V.5 from obscurity when he included the greater part of the readings of c (the siglum he himself retains) among the rich store of information presented in his work. F. Munari included the fragment in his catalogue, clearly pointing to the uncertainty surrounding its dating (‘aetatis incertae’, the same expression as used by Slater) while adding the information that it was not at that time to be found in Cesena. The succinct information offered by Slater is used by I. Marahrens for her dissertation and by W.S. Anderson for his edition. R.J. Tarrant, in turn, explicitly states that the fragment is now lost. In the following pages, however, I will attempt to show that this fragment is not lost, but still preserved in the Biblioteca Malatestiana in Cesena.
- Type
- Research Article
- Information
- Copyright
- Copyright © The Classical Association 2016
Footnotes
1
This study was carried out under the framework of Projects HUM-4534 and FFI2008-1843, which also financed a stay at the Biblioteca Malatestiana in Cesena in February 2014 for direct consultation of the codices Caesenates mentioned herein and facilitated access to other bibliography essential for the work. I would like to express my gratitude for the generous and efficient help offered by the staff of this library, and in particular by Mrs. Paola Errani, Head of its Conservation Service. Thanks are also due to Professor M.D. Reeve for reading a draft of this paper and for rechecking some of Heinsius's annotations. Once again I would like to thank three colleagues of mine for the disinterested help they have offered, namely Professors J.A. Estévez, J.A. Bellido and A. Sánchez. Finally, I wish to thank J.J. Zoltowski for the English translation.
References
2 The abbreviations used here are those of our Project for a critical-textual commentary on the Metamorphoses of Ovid, currently being carried out at the University of Huelva. They can be consulted, along with descriptions of the codices, at http://www.uhu.es/proyectovidio/pdf/descripcion.pdf. The essential works on Heinsius's manuscripts are still the two articles by M.D. Reeve, which in turn include the extremely important earlier bibliography: ‘Heinsius's manuscripts of Ovid’, RhM 117 (1974), 133–66Google Scholar (on the fragment which concerns us here, see especially 136 n. 9 and 153); ‘Heinsius's manuscripts of Ovid. A supplement’, RhM 119 (1976), 65–78 Google Scholar. See also Anderson, W.S., ‘Identification of another Heinsian manuscript’, CQ 26 (1976), 113–14CrossRefGoogle Scholar. More updated information, the fruit of the work of our Research Team and particularly of Professor J.A. Bellido, can be found at http://www.uhu.es/proyectovidio/esp/20_heinsiani.html J.A. Bellido has also published a comprehensive study of an analogous Heinsian manuscript on the text of Catullus: ‘Las notas a Catulo de A. Petreius y N. Heinsius (Berol. Diez. oct. 2474)’, ExClass 15 (2011), 123–200 Google Scholar.
3 Mentions of Langermann are constant, as can be seen in the corresponding entry in the index of P. Burman (ed.), Sylloge epistolarum a viris illustribus scriptarum (Leiden, 1727), 5.829. Langermann provided Heinsius with abundant epigraphical, numismatic and codicological material. See also Munari, F., ‘Manoscritti ovidiani di N. Heinsius’, SIFC 29 (1957), 98–114 Google Scholar, at 101 n. 1.
4 Reeve (n. 2 [1974]), 136 n. 9 and 153 n. 85. Oxon. Bodl. Auct. S.V.5 is in fact a copy, annotated in the margins, of the editio Gryphiana of 1563: Metamorphoses. Lugduni, apud Seb. Gryphium. See D.A. Slater, Towards a Text of the Metamorphosis of Ovid (Oxford, 1927), 12 and 24–8 for a description. Munari (n. 3), 99–100 includes it among the ‘Volumi con collazioni autografe di N. Heinsius’. The collation—taken to be that of Heinsius—was carried out in Strasbourg in 1653, according to Reeve (n. 2 [1974]), 136 n. 9.
5 P. Burmannus, Opera omnia, vol. 2: Metamorphoses (Amsterdam, 1727).
6 F. Munari, Catalogue of the MSS. of Ovid's Metamorphoses (London, 1957), no. 65, at 18.
7 I. Marahrens, Angefochtene Verse und Versgruppen in den Metamorphosen. Beiträge zu Ovids Sprache und Kompositionskunst (Diss., Heidelberg, 1971), 13. From Slater (n. 4), 12 she even adopts the mistaken statement that the fragment goes as far as 13.402 and not 403, a detail corrected by Slater himself at 24.
8 W.S. Anderson, P. Ovidius Naso. Metamorphoses (Munich and Leipzig, 2001), XVIII.
9 R.J. Tarrant, P. Ovidi Nasonis Metamorphoses (Oxford, 2004), ix. Both Anderson and Tarrant, in their respective descriptions of the contents of the fragment and in contrast with the more cautious ‘inerant’ of Slater (n. 4), introduce the verbal form ‘continebat’, which would appear to indicate that the lines pointed to there are its complete contents.
10 As read by Reeve (n. 2 [1974]), 153 n. 85, although he suspects that desinit should perhaps be read for deficit (this suspicion being confirmed per litt. in May 2014). J.A. Estévez, for his part, also reasonably opts for the reading ‘Ø c finit fragm. Caesen.’.
11 I quote from the edition of 1659: P. Ovidii Nasonis operum tomus II qui METAMORPHOSES complectitur. Nicolaus Heinsius, D. F. locis infinitis ex fide scriptorum exemplarium castigavit, & observationes adjecit. Amstelaedami, ex officinâ Elzevirianâ. A° M DC LIX. pp. 1–356, textus Metamorphoseos. Index. pp. 7–434, Heinsii notae. Addenda aut mutanda notis. Ad tomum II, p. 464. The note cited appears on p. 249. Burman (n. 5), on Met. 2.664 reproduces it identically.
12 This seems to be indicated by the editorial evidence. Though not in itself definitive proof, the register of consultations of the Biblioteca Malatestiana, which starts in 1877, seems to reinforce this suspicion.
13 Mentioned by Munari (n. 6), no. 63, at 18. The other is S.XXV.6 (s. XIV), our Cs3, also listed by Munari (n. 6), no. 64, at 18, although at no point did he handle either of the codices.
14 G.M. Muccioli, Catalogus codicum manuscriptorum Malatestianae Bibliothecae fratrum minorum conventualium (Cesena, 1784), 6–7 (http://www.malatestiana.it/cgi-bin/wxis.exe/?IsisScript=Opcat/window.xis&tag6666=muccioli/tomo2/006.jpg).
15 R. Zazzeri, Sui codici e libri a stampa della Biblioteca Malatestiana di Cesena. Ricerche e osservazioni (Cesena, 1887), 252 (http://www.malatestiana.it/cgi-bin/wxis.exe/?IsisScript=Opcat/window.xis&tag6666=zazzeri/252.jpg).
16 Munari (n. 6), no. 65, at 18; F.T. Coulson and B. Roy, Incipitarium Ovidianum: A Finding Guide for Texts in Latin Related to the Study of Ovid in the Middle Ages and Renaissance (Turnhout, 2000), no. 14, at 30; no. 178, at 68.
17 The description by R. Crociani is available only on the web: http://www.malatestiana.it/manoscritti/manus/S.1.5.pdf.
18 Cf. Crociani (n. 17): ‘i fogli di guardia I-II e I’-II’ moderni, non numerati, apposti in fase di restauro’. For the same reason less importance for our purposes can be attached to the note of sale which appears, with considerable crossing out, on the same page: Ego Thomas s[--]a[--] vendidi hunc librum ut constat manu Ser Nicolai Bartholomei.
19 Crociani (n. 17), n. 1.
20 I limit myself to these lines here because I am preparing the critical-textual commentary on Book 13 and I consider them to be a more than sufficient sample to demonstrate the identical nature of the two texts. Professors Bellido and Estévez have already collated the other books involved, this collation confirming the hypothesis I defend here, and they are preparing a detailed study on the contribution of the text of Cs. For a vindication of the quality of the text of this fragment, see G. Luck, Review of Tarrant (n. 9), ExClass 9 (2005), 249–71, at 253–5.
21 The first reading corresponds to the edition by Tarrant (n. 9). It should be understood that, as I am attempting here to confirm the identity of c and Cs (my reference Cs, in fact, is equivalent to c = Cs), I list only the readings pointed to by Langermann in S.V.5, dispensing with all the others I have gathered in the course of my own collation of the text and which I list in the critical apparatus of my commentary and which will also appear in the study by Bellido and Estévez. I mark with an asterisk those readings which appear to be exclusive to Cs at the present stage of my research, which encompasses the collation of more than 130 codices (ss. XI-XV). When a reading is attested in only one other MS apart from Cs, I will make a note to that effect.
22 Jahn, in his edition of 1828–1832 (P. Ovidii Nasonis quae supersunt Opera Omnia, 3 vols. [Leipzig, 1832], 2.2.xxiv), attests this reading in what he calls ‘cod. Rhen.’, the Rhenou(i)anus, but at least in Turicensis Bibl. Centr. 413 (Turicensis Bibl. Rhenouianus 46, s. XII), our Tu, identified with the Rhenouanus of E.C.Ch. Bach (see Munari [n. 6], no. 329, at 64), the reading is clearly igitur fictis.
23 Thus S.V.5, where the reading sû is attributed to c. Although this reading cannot be ruled out, neither can that of sñ (i.e. sine), which I am inclined to prefer.
24 I have found this spelling only in Mt: Matritensis MS 10038 Bibl. Nac. (olim Toletanus 102–8) (s. XII/XIII/XIV).
25 S.V.5 reflects the reading piscia, with a small square leaning on the second -i. However, I am inclined to suggest that this spelling was not supposed to represent i but l’ (= ul). I have also found the variant piscula attested in Vd: Vindobonensis Bibl. Nat. 207 (s. XII/XIII).
26 A reading attested only in So: Soriensis MS 4-H Bibl. Pub. (s. XIII2).
27 This is my correction of the reading attributed to c in S.V.5: uestrumque.
28 A variant documented only in Ab: Atrabaticus Bibl. Civ. 996 Caron (359 Catal. Génér.) (s. XII).
29 A variant documented only in B: Berolinensis Deutsche Staatsbibl. Diez B Sant. 9 (s. XII).
30 This is clearly the reading of Cs. However, Slater (n. 4), ad loc. seems to take it as dum when he states: ‘capit dum c. λαμβάνει ὡϛ ἄν αὐτῷ – ἕποιτο λαμβάνει ὡϛ ἄν αὐτῷ – ἕποιτο Plan.’, but that dum appears for cur in line 220, not here.
31 In MS Dr (Dresdensis Sächische Landesbibliothek, App. 1092, s. XII) the copyist also initially wrote pa- but corrected it immediately.
32 I can find this variant only in T7: Monacensis clm 29208(4 (antea 29007b), s. XII2, a codex with which Cs shares a good number of readings.
33 This is my correction of the reading published in S.V.5: talomonides.
34 The note in S.V.5 gives it as &phidonq 3 . The first element may be the result of a misinterpretation of the ý; the addition of the enclitic might stem from a misreading of the sign indicating the omission of the enclitic et which appears in most of the MSS. In any case, the reading of Cs is clear.
35 In S.V.5 it is given as C(h)roniumque, a possible reading, although I opt for the one given above.
36 Langermann reads it as aliumque. As can be seen, in the case of proper names the transcription is at times only approximate.
37 Thus Langermann, although the abbreviation (p'dañq 3 ) could also be read as predanumque.
38 Attested only in T7, apart from here.
39 That is, Cs coincides with the reading offered by Tarrant and by Heinsius before him (sed), as opposed to the variant modo which is accepted by the ed. Gryphiana of S.V.5. It is this divergence that is the reason for this reading of Cs being made explicit here.
40 The error is caused by the fact that the abbreviation of et is produced vertically downwards, in a sign similar to oʒ, but there is no doubt about its identification, as can be seen in the image (fig. 1):
Figure 1. (fol. 101v)
Besides, it is the sign used habitually throughout the fragment. There is another case of confusion on the part of Langermann caused by his reading of the same sign in 9.362, an error which logically reappears in Slater (n. 4), ad loc.
41 In S.V.5 it is given as accepto.
42 This indeed seems to be the reading, although uobis is not to be ruled out completely.
43 A reading found only in M2(a.c.): Marcianus Florentinus 223 (s. XI2).
44 That is, as opposed to the facit of the ed. Gryphiana/S.V.5.
45 That is, as opposed to the tanto of the ed. Gryphiana/S.V.5.
46 I have found this non-metrical reading attested only in Tu and the edition by Raph. Regius (Venice, 1493), although it is later found corrected in, for example, the Toscolano edition of 1526. This passage has the most striking discrepancy, as Langermann attributes rubefactum to c (‘tum c’). In fact, the divergence is again due to a misreading, since the manuscript gives rubefcâ, a spelling in which the tilde represents the abbreviation of –facta and is not the nasal tilde. In either case, the final vowel is clearly -a, not -u.
47 A variant attested only in Dr.
48 As opposed to the impositaque fero est of the ed. Gryphiana/S.V.5.
49 In the right-hand margin, in addition, in the space occupied by the scoring-out, the following words can still be made out, written perpendicular to the text by another, very different hand: Domine spes super om / super omnia bona sua (cf. Matt. 24:47).
50 Although it may seem obvious, it must be made clear that the copy from which Lucas Langermann took the readings he sent to Heinsius could never have been Cs2, as can be demonstrated by the two stretches of text which are duplicated by overlapping in the codex as we have it today (9.234-51 and 13.364-403). In those passages which are contained in S.V.5 and whose text differs between Cs and Cs2, it is naturally always the reading of Cs which is cited. These cases are: 9.235 silua Cs (sûma mg. scr. manus prima) : summam Cs2 • 9.244 mihi om. Cs : mihi Cs2 • 9.245 populis Cs : populi Cs2 • 9.250 uicit] uincit CsCs2 • 13.368 tantum Cs : tanto Cs2 • 13.371 proque] perque Cs : pro Cs2 • 13.381 fale Cs : fatale Cs2 • 13.383 re] tum Cs : tunc Cs2 • 13.386 uincit Cs : uicit Cs2 • 13.394 rubefacta Cs : rubefactaque Cs2 • 13.395 genuit uiridi Cs : uiridi genuit Cs2 • 13.396 Oebalio] eubalio Cs : ebalio Cs2 • 13.399 Hypsipyles] ýphiles Cs : ysiphiles Cs2 • post 13.399-401 transierat tandem portu uotoque potitus om. Cs : post 401 hab. Cs2 • 13.402 reuixit Cs : reuexit Cs2 • 13.403 imposita est sero Cs : impositaque est sero Cs2
51 For the internal layout of the two codices, see Crociani (n. 17). We should not rule out the possibility that Cs2 was simply mutilated to accommodate Cs (I owe this suggestion to the anonymous referee of this paper).
52 I owe this clear calculation to Prof. Estévez. Following this reasoning and always taking into account the margin of variability between the two figures, which is the result of the variability of the number of lines per page, we can hypothesize that the first of these lost quaternions—as far as fol. 72v of the original Cs2 (it should be remembered that the present numbering is modern, and certainly later than the amalgamation of Cs2+Cs)—went up to around 10.495 (c.1,040 lines); the second (as far as fol. 80v) up to the end of Book 11 (c.1,039 lines); the third up to 13.363 (991 lines). Obviously, we could also consider a more regular distribution, of 1,023 lines per quaternion (as far as 10.478 and 11.762 respectively for the first two quaternions, leaving 1,024 lines for the third).
53 It is trickier to hazard a guess at the reason for the coincidence, although Professor A. Ramírez de Verger suggests that it may well have something to do with censorship of the ‘unsuitable’ content of Book 9, which would explain both the deletion of the text from Cs2 and the permitted circulation, for ‘private study use’, of fragment Cs; compare, for example, the case of the contemporary MS Bernensis Bibl. Civ. 345 (Munari [n. 6], no. 36, at 13), in which fols 56–86 have been mutilated, making Books 8–10 illegible. However, it would still not be clear why the text that has been removed stretches almost to the halfway point of Book 13.
54 This implies, in turn, that at some point in its transmission fol. 82 was separate from the other folios of its quaternion (83–9), just as the first of the folios that make up the fragment (fol. 65) is a loose folio that closed a quaternion, as is demonstrated by the presence of a note in calce (on this, see below).
55 Here are some examples of variants not cited by Langermann, collected from Book 13. I omit minor graphical changes. Before the bracket I give first the reading of the ed. Gryphiana/S.V.5: 10 manu] magnu Cs(a.c.) • 35 illis] ille Cs • 47 siluestribus] siluestrisque Cs(a.c.) • 59 finxit] fixit Cs • 67 non haec] nec hec Cs • 86 fudi] fodi Cs • 99 rapta [sic Cs(a.c.)]] capta Cs(p.c.) • 100 nihil est] nil est Cs • 110 mundi] mondi Cs • 158 illa] arma Cs • 160 quam quae] quam cum (per comp.) Cs • 161 in promptu] in promptum Cs • 166 neque] nec Cs • 192 laudem ut] laudum Cs • 194 quo] quod Cs • 203] hoc] o Cs • 204 primaque] prima quoque (per comp.) Cs (a.c.) • 205 consilioque] consilique Cs (a.c.) • 214 simus] sumus Cs (a.c.) • 223 et] om. Cs • 228 disertum] desertum Cs • 233 impune] îpugne Cs || proteruus] proteruis Cs • 249 contentus] contêptus Cs • 260 fatisque] fatis Cs || Eunomon] eunemon Cs [Langermann, however, does note the variants of the other proper names in these lines] • 262 et] om. Cs • 264 diduxit] deduxit Cs • 268 refert] referam Cs • 276 oblitus regisque] oblitus regis Cs (p.c.) : oblitusque mei Cs (a.c.) • 286 uires] uiros Cs (a.c.) • 288 iccirco] îcirco Cs • 293 Pleiadasque] pleyadas Cs • 294 nitidumque] nudumque Cs • 348 nocte] om. Cs (a.c.) • 349 uici] uinci Cs • 350 uultuque] uultu Cs • 354 nisi] mihi (per comp. mi ) Cs (a.c. a manu prima) || pugnacem] pugnantem Cs • 378 praecipitique] praecipiti Cs • 384 qui ferrum] ferrum Cs • 387 an et] et an Cs (a.c.) || hunc sibi] hunc Cs || poscit] poscat Cs • 390 ne] nec Cs
56 Similar notes in calce to mark the boundary between quaternions are to be found in fols 65v, 73v, 81v, 89v. There is none on 99v, as this is not part of a quaternion but a bifolio. For the convenience of the reader, I recall here the structure of Cs: one loose folio (fol. 65) + four quaternions (fols 66–97) + one bifolio (fols 98–9) + one binion lacking the last folio (fols 100–2). As can be seen, its structure is much less regular than that of Cs2. We can even assume that this document must have circulated for some considerable time in the form of a bundle.
57 Another question still to be settled is whether the transfer of data from c (Cs) to S.V.5 pre- or post-dates that from AB (V2P2), since an initial analysis throws up contradictory results. Professor M.D. Reeve carried out a personal re-examination in May 2014 and concluded (per litt.) that ‘the order of the collations was AcB, but the symbols were used in the order c A+B’. Furthermore, he is convinced that the collation of B is in Heinsius's hand and that of c in Langermann's, while there is no sure evidence of who collated A. More details will be offered by Bellido and Estévez (n. 20).
58 See D. Savoia, ‘La Biblioteca Malatestiana’, in La casa dei libri. Dalla Libraria Domini alla Grande Malatestiana. Per i duecento anni della Biblioteca Comunale (1807–2007) (Cesena, 2007), 19–41, at 37 (although she persists in the error of considering the MS to be ‘degli inizi del XII secolo’). As regards the date of reception of the codex in the Biblioteca Malatestiana, Munari (n. 6), no. 63, at 18 erroneously gives ‘1597’. Again mistakenly, the date ‘1589’ is given by C. Ghiselli, ‘I restauri dei codici della Biblioteca Malatestiana dal 1950 al 2001’ (Diss., Univ. di Bologna, 2002), 44.
59 Crociani (n. 17) mentions two: ‘[Legatura di restauro …] Il restauro ha avuto luogo nel 1980 presso il Laboratorio di Santa Maria del Monte di Cesena, su un precedente intervento degli ultimi anni del secolo XIX ad opera del cosiddetto restauratore C.’ There is no documentary evidence of previous restorations. For the methods followed by the various restorers of the eighteenth to twentieth centuries, see Ghiselli (n. 58), 226–31, who, at 44–5, provides a description of the state of S.I.5 in particular.