The Last Poems of Ovid

by Ovid

The Manuscripts

The manuscript authority for the text of the fourth book of the Ex Ponto is significantly poorer than for the earlier books because of the absence of A, Hamburgensis scrin. 52 F. This ninth-century manuscript has been recognized since the time of Heinsius as the most important witness for the text of the Ex Ponto; it breaks off, however, at III ii 67.

The manuscript authorities for the fourth book can be placed in three categories. The fragmentary G is from a different tradition than the other manuscripts. B and C are closely related, and offer the best witness to the main tradition. The other manuscripts I have collated are more greatly affected by contamination and interpolation; of them M and F show some independence, while no subclassification can be made of H, I, L, or T.

G

The fragmentum Guelferbytanum, Cod. Guelf. 13.11 Aug. 4°, generally dated to the fifth or sixth century, is the oldest manuscript witness to any of Ovid's poems. Part of the collection of the Herzog August Bibliothek in Wolfenbüttel, it was discovered by Carl Schoenemann, who published his discovery in 1829; details of his monograph will be found in the bibliography. The two pieces of parchment are a palimpsest, having been reused in the eighth century for a text of Augustine; later they were incorporated into a bookbinding. As a result of this treatment they are in extremely poor condition.

G contains all or part of ix 101-8, ix 127-33, xii 15-19, and xii 41-44. To make it perfectly clear when G is a witness to the text, I have not grouped it with other manuscripts, but have always specified it by name. If G is not mentioned in an apparatus entry, it is not extant for the text concerned.

G is written in uncial script, with no division between words but with indentation of the pentameters. Its one contribution to the establishment of the text is at ix 103, where it reads quamquam ... sit instead of the more usual quamquam ... est found in the other manuscripts. In general, the text offered by G is surprisingly poor. At ix 108 it reads fato for facto, at ix 130 it has the false and unmetrical spelling praeces, at ix 132 it has misscelite for misi caelite, at xii 17 it reads lati for dilati, and at xii 19 naia for nota. These errors demonstrate that the rest of the tradition does not descend from G.

Korn gives an accurate transcription of the fragment in the introduction to his edition; photographs of parts of the fragment can be found at Chatelain, Paléographie des classiques latins, tab. xcix, 2 and E. A. Lowe, Codices Latini Antiquiores, vol. IX, p. 40, no. 1377.

B and C

Monacensis latinus 384 and Mon. lat. 19476, both dated by editors to the twelfth century, are descended from a common ancestor. This is easily demonstrated by the large number of shared errors not found in other manuscripts[9]. At iv 36 B and C have intendunt for the correct intendent, at viii 6 uolo for uoco, at viii 18 perueniemus for inueniemur (-ntur,-mus), at viii 44 illa for ulla, at viii 89 cara for care, at ix 44 fingit for finget, at ix 71 quod for cum (FILT) and ut (HM), at ix 92 praestat for perstat, at ix 97 et for ut, at xiii 5 certe est for certe, and at xiv 30 culpatus for culpatis. In some of these passages B's still visible original reading has been corrected by a later hand. In other passages it is clear from the signs of correction that B originally agreed with C in distinctive readings now preserved in C alone: subito for sed et (iii 27), erat for eras (vi 9), occidit for occidis (vi 11), suspicit for suscipit (ix 90), parent for darent (xvi 31).

B and C on the whole offer a better text than any other manuscript. At iii 44 B¹ and C omit the lost pentameter, where the other manuscripts offer interpolations. At iv 11 they alone give the probably correct solus for tristis, at xii 3 aut for ast, and at xvi 31 tyrannis (conjectured by Heinsius) for tyranni. At v 40 C and B² alone have the correct mancipii ... tui for mancipium ... tuum.

Both manuscripts naturally have readings peculiar to themselves. B has about fifty unique readings. It places iii 11-12 after 13-14, omits v 37-40, and interchanges viii 49-50 and 51-52. At iv 34 B alone has erunt (for erit), conjectured by Heinsius; C omits the word. Similarly, at xi 21 B and F¹ have mihi, omitted by C; the other manuscripts have tibi. B has ab at i 9 for the other manuscripts' in; ab is possibly the true reading.

Under the influence of Ehwald, modern editors have wrongly taken some of B's other readings to be correct, placing aspicerem in the text for prospicerem at ix 23, ara for ora at ix 115, and illi for illum at ix 126. At ix 73 editors print B and T's quem, which is clearly an interpolation for the awkward transmitted reading qua.

Unlike C, B has been quite heavily corrected by later hands.

C has more than one hundred readings peculiar to itself. Two of them I have accepted as correct: summo (for summum; H has mundum) at iii 32, and horas (that is, oras) at vii 1; the reading is also given by I. It is possible that C's correptior should be read at xii 13 for correptius. At xiv 38 C's sceptius is the manuscript reading closest to the correct Scepsius restored by Scaliger.

Most of C's errors are trivial, but at some points it departs widely from the usual text. It omits ix 47 and xiv 37, and interchanges the second hemistichs of iii 26 and 28; xvi 30 is inserted by a later hand, perhaps in an erasure. At viii 43 it has in uita for officio, at xiii 12 contra uiam for nouimus, at xiv 36 in for loci, and at xv 31 colloquio for uerum quid.

C also contains a greater number of purely palaeographical errors than any other manuscript: hunc for nunc (i 25), humeris for numeris (ii 30), hec for nec (ix 30), lucos for sucos (x 19), hasto for horto (xv 7), ueiiuolique for ueliuolique (xvi 21), pretia for pr(o)elia (xvi 23).

B and C sporadically offer the third declension accusative plural ending -is (ix 4 fascis C, ix 7 partis C, ix 73 rudentis B, x 17 cantantis B, xii 30 albentis B). But more usually all manuscripts, including B and C, have the accusative in -es: compare for example ii 27 partes, iii 53 purgantes, ix 35 praesentes, and ix 42 fasces. The manuscripts show a similar variation in the earlier books of the Ex Ponto. The ninth-century Hamburg manuscript (A) sometimes offers accusatives in -is where the other manuscripts, even B and C, have -es (I iv 23 partis, I v 11 talis, I vi 39 ligantis, I vi 51 turris). At I ii 4, A has omnes, where C¹ has omnis, and in general even in A the accusative in -es is the predominant form. For example, A offers auris at II iv 13 and II ix 25, but aures at I ii 127, I ix 5, II v 33, and II ix 3. In view of the instability of the manuscript evidence[10], I have normalized the ending to -es in all cases, considering the instances of -is to be scribal interpolations.

Similarly, I have used the form penna at iv 12 and vii 37, where C offers pinna. Penna is the form given in the ancient manuscripts of Virgil, and attested by Quintilian.

MFHILT

The other manuscripts I have collated belong to the vulgate class. They are not related to each other in the sense that B and C are related, nor does any of them possess independent authority as does G. Within the group firm lines of affiliation are hard to establish, and each of the manuscripts attests a handful of good readings that are found in few or none of the others, either by happy conjecture, or because a reading that was in circulation at the time as a variant chanced to get copied into a few surviving manuscripts. Professor R. J. Tarrant has noted that the presence of the Ex Ponto in north-central France 'can be traced from the eleventh century onwards, first from echoes in Hildebert of Lavardin and Baudri de Bourgeuil, later from the extracts in the Florilegium Gallicum, and finally from the complete texts [which include our H and F] ... that emanate from this region toward the end of the twelfth century' (Texts and Transmission 263); the vulgate manuscripts seem to have been propagated from the text current in the region of Orléans.

M and F show some originality. Their readings at xvi 33 differ somewhat from the version of that passage in HILT. F¹'s interpolation for the missing pentameter at iii 44 differs from that of MHILT, while M has an interpolated distich following x 6 that is not otherwise attested.

Of the other manuscripts, I agrees with C in reading horas (=oras) for undas at vii 1, while T is the only manuscript collated to have the correct laeuus at ix 119 in the original hand (F² gives it as a variant reading). Similarly, H and L each have a few peculiar variants.

As a group MFHILT offer a good picture of the readings current in the later mediaeval period, and only rarely have I been obliged to cite a vulgate manuscript from the editions of Heinsius, Burman, or Lenz as testimony for a variant.

M

Heinsius did not have knowledge of B or C, and seems to have considered his codex Moreti (preserved at the Museum Plantin-Moretus in Antwerp as 'Latin, n° 68 [anc. 43] [salle des reliures, n° 32]' in Denucé's catalogue of the museum's collection) to be the best of the poor selection of manuscripts available; at xvi 33, understandably despairing of restoring the true reading, he accepted M's reading pending the discovery of better manuscripts.

M was dated by Heinsius to the twelfth or thirteenth century; Denucé assigns it to the twelfth century.

At viii 85 M alone has the correct ullo for the other manuscripts' illo; this could naturally have been recovered by conjecture. At x 1 it has cumerio, the closest reading in the manuscripts collated to the correct Cimmerio; but Professor R. J. Tarrant informs me that Cimmerio is also found in British Library Harley 2607.

M has suffered from a certain degree of interpolation. Following x 6 there is the spurious distich set cum nostra malis uexentur corpora multis / aspera non possum perpetiendo mori. At ii 9 Falerno is a deliberate alteration of Falerna. At x 49 Niphates is an interpolation from Lucan III 245. At xiii 47 duorum (also given as a variant reading by F²) looks like an attempt to correct the cryptic transmitted reading deorum, and at xv 15 tellus regnata is presumably a metrical correction following the loss of -que from regnataque terra, the reading of the other manuscripts. At xvi 25 eticiusque looks to be a deliberate alteration of Trinacriusque, but I am not sure what the interpolation means.

F

Francofurtanus Barth 110, used by Burman, shows some signs of independence. At iii 44, where a pentameter has been lost, B and C omit the line, while the other manuscripts, including M, have the interpolation indigus effectus omnibus ipse magis; F has the separate interpolation Achillas Pharius abstulit ense caput, also found in Heinsius' fragmentum Louaniense. F omits viii 51-54, at xi 1 reads Pollio for Gallio, and at xvi 33 has a reading somewhat different from those offered by the other manuscripts.

F alone of the manuscripts collated offers the correct audisse (for audire) at x 17. At xi 21 it and B alone have the correct mihi for tibi (omitted by C). At xiv 7 it has the probably correct muter for mittar, also found in Bodleianus Canon. lat. 1 and Barberinus lat. 26, both of the thirteenth century. With the exception of muter, these readings could have been recovered by conjecture; given the separative interpolation at iii 44, F differs surprisingly little from the other manuscripts.

H

The thirteenth-century Holkhamicus 322, now British Library add. 49368, contains (with I) the correct hanc at i 16, the other manuscripts having ha, ah (B), or a (C). At xvi 30, where I have printed leuis, the reading of most manuscripts, H has leui, the conjecture of Heinsius; Professor R. J. Tarrant informs me that the same reading is found in Othob. lat. 1469. At iv 45 H's qua libet is the manuscript reading closest to Heinsius' correct quamlibet; most manuscripts have quod licet.

Most other variants in H are trivial errors, although there seems to have been deliberate scribal alteration at x 18 (sucus amarus erat for lotos amara fuit), xiv 38 (Celsius for the usual Septius; Scaliger restored Scepsius), xvi 3 (ueniet for uenit et; presumably the intermediate step was uenit), and perhaps at xiv 31 (miserabilis for uitabilis).

I

The thirteenth-century Laurentianus 36 32, Lenz's and André's m, has the correct perstas at x 83 for praestas; its reading is also found in P and as a variant of F². At vii 1 it shares with C the reading horas (=oras), which I have printed in preference to the usual undas.

At viii 15 I has the hypercorrect nil for nihil, and at xiii 26 ethereos ... deos for aetherias ... domos, but in general has few signs of deliberate alteration.

L

Lipsiensis bibl. ciu. Rep. I 2° 7, of the thirteenth century, has haec at ix 103 for the other manuscripts' et. Haec restores sense to the passage, and was the preferred reading of Heinsius; I consider it a scribal conjecture, now rendered obsolete by Professor R. J. Tarrant's more elegant quae. L's text has clearly been tampered with at xiv 41 (populum ... uertit in iram for populi ... concitat iram), but in general seems to have suffered little from interpolation. It is, however, of little independent value as a witness to the text.

T

Turonensis 879, written around the year 1200, was first fully collated by André for his edition; Lenz had earlier reported its readings for IV xvi and part of I i. At ix 119 only T and F² of the manuscripts collated have the correct laeuus, although other manuscripts come close, and the reading could have been recovered by conjecture. At xv 40 T reads transierit saeuos for transit nostra feros; clearly nostra was at some point lost from the text, and metre forcibly restored.

P

I have also collated the thirteenth-century Parisinus lat. 7993, Heinsius' codex Regius. At ix 46 P offers the correct cernet for credet; cernet is also the reading of M after correction by a later hand and of the thirteenth-century Gothanus membr. II 121. At vi 7 P alone of collated manuscripts agrees with C in reading praestat for the correct perstat. P agrees with L in reading niuibus for the other manuscripts' nubibus at v 5, adeptum for ademptum at vi 49, signare for signate at xv 11, and in the orthography puplicus for publicus at ix 48, ix 102, xiii 5, and xiv 16. The manuscript has many corruptions: a few examples are i 30 igne for imbre, ii 18 supremo for suppresso, iv 6 pace for parte, vi 34 uirtus for uirus, vii 15 piacula for pericula, ix 42 praeterea for praetextam, x 63 in harena for marina, xiv 39 conuiuia for conuicia, and xvi 24 sacri for scripti. However, P has no unique variants with any probability of correctness. To have given a full report of P would have involved a considerable expansion of an already long apparatus, and I have cited the manuscript only occasionally, where a reading is only weakly attested by the other manuscripts.