Hostname: page-component-cd9895bd7-7cvxr Total loading time: 0 Render date: 2024-12-23T17:52:27.946Z Has data issue: false hasContentIssue false
  • English
  • Français

CRITICAL NOTES ON SERVIUS’ COMMENTARY ON VIRGIL (SERV. ON AEN. 11.741; ECL. 2.58; ECL. 4.4)

Published online by Cambridge University Press:  14 March 2023

Stefano Poletti
Stefano Poletti*
Affiliation:
Albert-Ludwigs-Universität Freiburg
*
[email protected]
Rights & Permissions [Opens in a new window]

Abstract

This article discusses three textual problems in Servius’ commentary on Virgil (Serv. on Aen. 11.741; Ecl. 2.58; Ecl. 4.4). In two notes a new conjecture is proposed; in one passage a transmitted reading, so far neglected by earlier editors, is supported.

Keywords

ServiusVirgilian exegesistextual criticismmanuscript traditions
Type
Research Article
Information
The Classical Quarterly , Volume 72 , Issue 2 , December 2022 , pp. 868 - 876
Check for updates
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of The Classical Association

This article discusses three textual and interpretative problems in Servius’ commentary on Virgil (Serv. on Aen. 11.741; Ecl. 2.58; Ecl. 4.4).Footnote 1 In two notes (Ecl. 2.58; Ecl. 4.4) a new conjecture is proposed. In the case of the scholia to Aen. 11.741 and Ecl. 2.58, a new evaluation of the manuscript tradition would give us reason to adopt a different text from what has been printed by editors thus far. Furthermore, the textual problem of Serv. on Ecl. 2.58 will be contextualized in a discussion of the consistency of the scholium, which is apparently the result of a conflation of different scholia.

1. SERV. ON AEN. 11.741: TARCHON MORITVRVS

Verg. Aen. 11.741–2 haec effatus equum in medios moriturus et ipse | concitat et Venulo aduersum se turbidus infert.

So saying, he [sc. Tarchon] spurs his horse into the throng, ready himself also to die, and charges like a whirlwind full at Venulus.

As stated by Horsfall, ‘in contrast with V(irgil)'s common use of moriturus […], periturus […], T(archon) is not about to die’.Footnote 2 This point was already made by Servius, who comments (Murgia's text and apparatus criticus):

MORITVRVS ET IPSE moriturus animo, nam moriturus non est. quod autem ait ‘et ipse’, aut ad Camillam aut ad Venulum respicit.

moriturus (animo)] moriturus F Pc γ  morituri J Q N [Ʃ]  moritur θ  morituro W U  moriturus in Pa

The problem at stake here is the subjective or objective interpretation of moriturus (‘determined/ready to die’, ‘certain of dying’, etc. as opposed to ‘about/doomed to die’).Footnote 3 Servius also deals with the very same topic in another scholium, on Coroebus periturus in Book 2 (a passage where the meaning of the participle was debated: melior sensus est), and quotes Tarchon's instance as a parallel:Footnote 4

Verg. Aen. 2.407–8 non tulit hanc speciem furiata mente Coroebus | et sese medium iniecit periturus in agmen […]

Serv. PERITVRVS melior sensus est, si ad dimicantis referatur affectum: sicut de Tarchonte, de quo dixit [11.741] ‘et medios fertur moriturus in hostes’,Footnote 5 cum uicerit.

At the beginning of the scholium to Aen. 11.741, all editors print moriturus animo, apparently taking animo as an ablative of respect (‘ready to die in his soul’). But other instances of animo as an ablative of respect show that the syntagm moriturus animo is unexpected and suspect.Footnote 6 Equally suspect is the repetition of moriturus (MORITVRVS ET IPSE moriturus animo […]). Although Murgia's apparatus criticus indicates that morituri is the reading of the Servian archetype ([Ʃ]), he rejected this reading and followed previous editors in accepting moriturus found in the auctus-witness (F) and in some manuscripts of Servius.Footnote 7 But does morituri animo really not make any sense?

MORITVRVS ET IPSE morituri animo: nam moriturus non est.

Given the context, morituri must be a genitive singular depending on animo, and the clear meaning of morituri animo effectively glosses moriturus: ‘(he spurs his horse/performs this action) with the mindset of a moriturus, of someone who is going/ready/doomed to die’. The syntax of morituri might seem rather elliptic, whereas moriturus straightforwardly resumes the lemma at the beginning of the scholium. But this is precisely the reason why morituri should be considered a lectio difficilior: given the lemma and the subsequent moriturus non est, the easy corruption of morituri to moriturus could well have occurred independently in F, in Pc and in γ. Moreover, morituri animo seems to correspond to the structure ‘affectus + genitive-singular present participle’, which Servius sometimes uses to define a character's state of mind, their own ‘subjective perspective’, as in the case of Coroebus’ dimicantis affectus (see above).Footnote 8 In the Servian corpus there is no further instance of animus + future participle, but this is not a serious obstacle. The structure is quite common in Tiberius Claudius Donatus and in Donatus’ commentary on Terence, where animo + genitive-singular participle indicates the mindset with which a character performs an action.Footnote 9

The text printed so far, though suspect, is perhaps not unacceptable. However, Murgia's new evaluation of the manuscript tradition invites us to consider seriously the genitive morituri.

2. SERV. ON ECL. 2.58: CORYDON'S IMAGINATION

In Verg. Ecl. 2.56–61 Corydon is alone, lamenting his unrequited love for Alexis: he talks to himself (‘rusticus es …’) and then addresses his beloved (‘quem fugis, a, demens?’).

‘rusticus es, Corydon; nec munera curat Alexis,
nec, si muneribus certes, concedat Iollas.
heu heu, quid uolui misero mihi? floribus Austrum
perditus et liquidis immisi fontibus apros.
quem fugis, a, demens? habitarunt di quoque siluas           60
Dardaniusque Paris.’

‘Corydon, you are a clown! Alexis cares naught for gifts, nor if with gifts you were to vie, would Iollas yield. Alas, alas! What hope, poor fool, has been mine? Madman, I have let in the south wind to my flowers, and boars to my crystal springs! Ah, idiot, whom do you flee? Even the gods have dwelt in the woods, and Dardan Paris.’

Here is Servius’ commentary on line 58 as printed in Thilo's edition (the apparatus criticus is based on a new collation of the witnesses).Footnote 10

Serv. on Ecl. 2.58 HEV HEV QVID VOLVI MISERO MIHI quomodo eum dicit discedere, quem supra cum eo diximus non fuisse? Nam ait [Ecl. 2.4–5] ‘solus montibus et siluis’. sed ratione non caret: Epicurei enim dicunt, quod etiam Cicero tractat [Cic. Tusc. 5.96], geminam esse uoluptatem, unam quae percipitur, et alteram imaginariam, scilicet eam quae nascitur ex cogitatione. unde ita debemus accipere, hunc usum per cogitationem illa imaginaria uoluptate, qua et cernere et adloqui uidebatur absentem. sed postquam obiurgatione sua in naturalem prudentiam est reuersus, caruit utique illa imaginaria uoluptate, ubi nunc sibi se offuisse dicit per hanc ratiocinationem: [Ecl. 2.56–7] ‘rusticus es, Corydon: nec munera curat Alexis, nec, si m. c. c. Iollas’.

quomodo] quod Y | eum dicit … ratione non om. Le Pc | dicit discedere] discedere dicit E Pb Y Scpc | nam] nam quod Bc M Pb σ (Guarinus, Masvicius, Burman, Lion) | ratione] sed ratione Bo H (Cennini, Stephanus, Fabricius, Thilo) : ratione cett.

ALAS, ALAS! WHAT HOPE, POOR FOOL, HAS BEEN MINE? How can he [sc. Corydon] say that he [sc. Alexis] is going away, who [sc. Alexis], as we said above, was not with him [sc. Corydon]? In fact, the poet says [of Corydon] [Ecl. 2.4–5] ‘alone with the mountains and forests’. But there is an explanation. For the Epicureans say, as is also discussed by Cicero [Cic. Tusc. 5.96], that pleasure comes in two kinds: one that is due to perception and another due to imagination, which is born from our thinking. Therefore, we must understand that Corydon has experienced this imagined pleasure, thanks to which he seemed to see his absent beloved and talk to him. But after his own reproach had led him to regain his natural clarity of thought, he was certainly deprived of that imagined pleasure, as here he says he has hurt himself by this reasoning: [Ecl. 2.56–7] ‘Corydon, you are a clown! Alexis cares naught for gifts, nor if with gifts you were to vie, would Iollas yield.’

Two surrounding scholia are related to this one: the scholium to lines 56–7 on the self-reproach (Serv. on Ecl. 2.56 RVSTICVS ES CORYDON arguit se stultitiae, quod eum se sperat placare muneribus, qui potest habere meliora, nam supra ait ‘delicias domini’) and the scholium to line 60 on Corydon's phantasia (QVEM FVGIS A. D. iterum per phantasiam quasi ad praesentem loquitur. […]).

In the scholium to Ecl. 2.58 Servius first points out a seeming contradiction: Corydon says that Alexis is going away, while at the beginning of the Eclogue Corydon is presented as being alone.Footnote 11 Then the commentator explains that the character is using his imaginatio Footnote 12 and that, after the self-reproach of lines 56–7, he comes back to his ‘clarity of thought’. The structure of the scholium is very odd indeed. The opening remark (eum dicit discedere) refers not to the lemma ‘heu heu, quid uolui misero mihi?’ but to line 60 ‘quem fugis?’ without any indication of a cross-reference (for example paulo post).Footnote 13 Only at the end do we find a (relatively) clear reference to the lemma (nunc sibi se offuisse dicit: contrast eum dicit discedere) and its context (56–7).Footnote 14 Apparently some exegetic material related to line 60 was conflated with that relating to lines 56–8, but a transposition of the scholium of line 58 (or also part of it) to line 60 is not possible, since at Serv. on Ecl. 2.60 there is already a note on phantasia (see above) and ubi nunc … dicit anchors the scholium to line 58 (that is, to the context of Corydon's ‘rationality’). Hence the most probable hypothesis is that Servius is merging scholia related to different lines in a rather mechanical and clumsy way (as he often does),Footnote 15 with the aim of offering a general reflection on this pivotal passage (56–60).Footnote 16 In the scholium to line 60, he then presupposes this detailed note on imaginatio (iterum = ‘again’, that is, after a moment of naturalis prudentia).

The textual problem at the opening of Serv. on Ecl. 2.58 gives further food for thought about this issue. Here is the text transmitted by the majority of our witnesses:

quomodo eum dicit discedere, quem supra cum eo diximus non fuisse? nam ait ‘solus montibus et siluis’. ratione non caret: Epicurei enim …

or, with a different punctuation,

quomodo eum dicit discedere, quem supra cum eo diximus non fuisse (nam ait ‘solus montibus et siluis’)? ratione non caret: Epicurei enim …

In the Servian corpus (especially in the auctus) there are many instances of quaestiones introduced by quomodo on possible contradictions in the Virgilian text.Footnote 17 The addition of sed (sed ratione non caret), found in MSS Bo H and printed so far in many editions, is not strictly necessary, but highlights very well the editors’ discomfort with the conciseness of ratione non caret.

Some manuscripts offer a quite different scenario:

Bc M Pb σ: quomodo eum dicit discedere (discedere dicit Pb),Footnote 18 quem supra cum eo diximus non fuisse? nam quod ait ‘solus montibus et siluis’ ratione non caret: Epicurei enim …

Y: quod eum discedere dicit, quem supra cum eo diximus non fuisse (nam ait ‘solus montibus et siluis’),Footnote 19 ratione non caret: Epicurei enim …

These readings, though perhaps conjectural, deserve serious consideration.Footnote 20 Servius uses this expression (quod dicit/ait … ratione non caret, ‘the fact that he says … has a reason’) to clarify seeming inconsistencies in the Virgilian text.Footnote 21 In this scholium the quod-clause would provide a subject for ratione non caret, which otherwise would be hanging in the air—hence also the sed added by many editors. While the reading of MSS Bc M Pb σ is not convincing (nam must introduce the quotation solus montibus et siluis, certainly not the answer to the quaestio itself), the reading of MS Y makes perfect sense and is very likely to be the right reading, which was then corrupted in the typical quomodo opening many scholia. Following MS Y, I would go a step further, by offering a correction that could better account for the quomodo of the other witnesses: quod modo (‘the fact that now he says’, in contrast with supra diximus … nam ait …).Footnote 22 Servius uses quite often the expression quod modo, also in comparisons of two (seemingly) contradictory passages.Footnote 23 Obviously, the presence of modo at the beginning of the scholium to Ecl. 2.58 seems to clash with ubi nunc … dicit at the end of it: quod modo dicit would perfectly fit with quem fugis? as a lemma, while it is less suitable to introduce a cross-reference to a nearby passage. But if, as demonstrated above, a scholium from line 60 was conflated by Servius quite mechanically here, quod modo could well be the original reading. This is a further element to consider in the issue of the odd structure of the scholium to line 58.

3. SERV. ON ECL. 4.4: THE CUMAEAN SIBYL

I shall concentrate on the Servian gloss to the famous Cymaeum carmen in the fourth Eclogue (Verg. Ecl. 4.4–5):

ultima Cymaei uenit iam carminis aetas;
magnus ab integro saeclorum nascitur ordo.

Now is come the last Age of Cumaean song; the great line of the centuries begins anew.

Here is Servius’ scholium (Thilo's edition):

Serv. on Ecl. 4.4 VLTIMA CYMAEI V. I. C. A. Sibyllini, quae Cumana fuit et saecula per metalla diuisit, dixit etiam quis quo saeculo imperaret, et Solem ultimum, id est decimum uoluit: nouimus autem eundem esse Apollinem, unde dicit [4.10] ‘tuus iam regnat Apollo’. dixit etiam, finitis omnibus saeculis rursus eadem innouari: quam rem etiam philosophi hac disputatione colligunt, dicentes, completo magno anno omnia sidera in ortus suos redire et ferri rursus eodem motu. quod si est idem siderum motus, necesse est ut omnia quae fuerunt habeant iterationem: uniuersa enim ex astrorum motu pendere manifestum est. hoc secutus Vergilius dicit reuerti aurea saecula et iterari omnia quae fuerunt.

dixit etiam quis … et iterari omnia quae fuerunt om. L (= auctus-witness)

Servius is speaking of the prophecy of the Sibyl and the myth of the metals and the Ages,Footnote 24 which is alluded to also in the scholium to Ecl. 4.10 (mentioned in Serv. on Ecl. 4.4): TVVS IAM R. APOLLO et ultimum saeculum ostendit, quod Sibylla Solis esse memorauit […]. Serv. on Ecl. 4.4 is a well-known text, which is often quoted, following Thilo's (dubious) textual arrangement, in many studies on the fourth Eclogue as well as on Sibylline literature.Footnote 25 The phrasing of the scholium is in itself pretty flat and repetitive (see the repetition of dixit etiam), but the weird expression Sibyllini quae does deserve attention. Thilo prints the paradosis:

VLTIMA CYMAEI V. I. C. A. Sibyllini, quae Cumana fuit et saecula per metalla diuisit, dixit etiam quis quo saeculo imperaret […]

(CUMAEAN) Sibylline, the one [Sibyl] who was from Cuma and divided the Ages by metals, also prophesied who would rule in which Age […]Footnote 26

The relative pronoun quae clearly refers to an implicit Sibylla, but the formulation is particularly awkward and hardly acceptable. If we consider ancient Virgilian exegesis besides Servius, two elements stand out. From an exegetical point of view, Servius is implicitly rejecting the interpretation of Cumaei as a reference to the Asian Cyme and Hesiod as a source of the myth of the Ages (saecula per metalla diuisit),Footnote 27 while also hinting at the existence of different Sibyllae defined on a geographical basis (quae Cumana fuit).Footnote 28 From a textual point of view, we have a confirmation that a Sibylla before quae is needed.Footnote 29 The following corrections were proposed:

  • • Thilo in app. crit.: Sibyllae, quae Cumana fuit et saecula per metalla diuisit. dixit etiam …; and: Sibyllini quod Cumanae fuit quae saecula per metalla diuisit. dixit etiam …Footnote 30

  • • Guarinus: Sibyllini, <id est Sibyllae>, quae Cumana fuit et saecula per metalla diuisit. dixit etiam …Footnote 31

  • • Corssen: Sibylla, quae Cumana fuit et saecula per metalla diuisit, dixit etiam …Footnote 32

As noted by Guarinus, both the meaningful gloss CYMAEI Sibyllini and the transmitted quae should be preserved.Footnote 33 Guarinus's solution is interesting for two other reasons:

  • • dixit etiam is also found at the beginning of the sentence some lines below (dixit etiam finitis, etc.);

  • • the interpretation of Sibyllini … diuisit as an ‘autonomous’ sentence is supported by the auctus-witness (Leid. Voss. Lat. O. 80), which omits the rest of the note after diuisit and provides a text that can be understood only if a genitive Sibyllae is added.Footnote 34

Nevertheless, if we keep the adjective Sibyllini, any solution involving the genitive Sibyllae is less effective, since the transmitted quae most likely refers to the subject of dixit (see the syntax in Corssen's solution). Hence the most attractive solution reads as follows:

Sibyllini. <Sibylla>, quae Cumana fuit et saecula per metalla diuisit, dixit etiam … Footnote 35

Sibylla was ‘absorbed’ by the preceding Sibyllini by a sort of haplography.Footnote 36

Footnotes

Parts of this paper were presented at the research seminars organized by Fabio Stok at the University of Roma Tor Vergata and at the Latin Grammarians Forum (Trinity College Dublin, 30–31 May 2019) organized by Anna Chahoud and Elena Spangenberg Yanes: I thank all the participants for their comments. Dániel Kiss, Giuseppe Ramires, Alessandro Russo, Ernesto Stagni and Fabio Stok read earlier versions of this paper: I thank them all for their useful comments. Finally, I wish to thank Bruce Gibson and the anonymous referee of CQ for their valuable suggestions.

References

1 For Serv. on Aen. 11.741 we now have Murgia, C.E. and Kaster, R.A. (edd.), Serviani in Vergili Aeneidos libros IX–XII commentarii (Oxford, 2018)Google Scholar. For Serv. on Ecl. 2.58 and 4.4 we must still rely on Thilo, G. (ed.), Servii grammatici qui feruntur in Vergilii Bucolica et Georgica commentarii (Leipzig, 1887)Google Scholar. On Servius’ manuscript tradition, see Murgia, C.E., Prolegomena to Servius 5 — The Manuscripts (Berkeley, 1975)Google Scholar, Murgia and Kaster (this note), xi–xxxviii. On Servius’ commentary, see Zetzel, J.E.G., Critics, Compilers, and Commentators (Oxford, 2018), 262–3Google Scholar. Where not otherwise specified, I quote Servius from Thilo (using italics only for the auctus-text), Virgil from Conte, G.B. and Ottaviano, S., P. Vergilius Maro Bucolica Georgica (Berlin and Boston, 2013)Google Scholar and Conte, G.B., P. Vergilius Maro Aeneis (Berlin and Boston, 2019 2)Google Scholar. Translations of Servius are mine. For Virgil I use the translation of Fairclough, H.R. and Goold, G.P., Virgil (Cambridge, MA, 1999–2000)Google Scholar.

2 Horsfall, N., Virgil, Aeneid 11. A Commentary (Leiden and Boston, 2003)CrossRefGoogle Scholar, ad loc.

3 On the (often debated) meaning of moriturus/periturus in Virgil, see now Grotto, F., ‘Frustra mori: per l'esegesi di Verg. Aen. 4, 415 e Stat. Theb. 9, 726–727’, MD 84 (2020), 173–96Google Scholar.

4 Servius hints at the ‘subjective’ connotation of moriturus/periturus also when commenting on the words of a character who defines himself/herself as periturus/moriturus (Serv. on Aen. 4.642 […] nam moritura nihil timebat, cf. Verg. Aen. 4.604 quem metui moritura?; Serv. on Aen. 10.881), but he recognizes that moriturus/periturus has usually an ‘objective’ meaning: Serv. on Aen. 10.341 quando dico ‘moriturus est’, uere moriturus est; moribundus autem non uere, sed similis morienti est (cf. also Serv. on G. 4.457, 10.501; Serv. auctus on Aen. 8.583).

5 The quotation is imprecise. At Verg. Aen. 11.741 the Virgilian manuscripts read in medios moriturus et ipse (et ipse] in hostis γ) and the Servian scholium ad loc. also discusses et ipse (see above). Other similar passages led to the confusion: Aen. 2.511, 9.400, 9.554 (cf. Conte [n. 1], ad loc.).

6 See e.g. Serv. on Aen. 1.613 OBSTIPVIT animo perculsa est; Serv. auctus on Aen. 10.858 HAVD DEIECTVS non deiectus animo.

7 I agree with Murgia (n. 1) on the archetypal status of morituri, which is transmitted in three witnesses belonging to different families. In Murgia's reconstruction, the moriturus of some Servian manuscripts is either a conjectural reading or a fortunate mistake or a contamination from Servius auctus. Most likely, the reading of Servius auctus persuaded Murgia to print moriturus. The other transmitted readings are unimportant, but they do suggest that the ending of the participle was ‘unstable’ in the manuscript tradition.

8 Cf. Serv. on Aen. 9.445 [sc. ‘placida morte’] ex adfectu pereuntis dictum est. See also animus + genitive, typical of Servius auctus to express the point of view of a character: Serv. auctus on Aen. 1.464 [sc. ‘inani’] ad stupentis animum rettulit (but animus is here in the Virgilian verse); 4.141 ex animo Didonis; 9.426; 12.636. On focalization in Virgilian exegesis, see Rosati, G., ‘Punto di vista narrativo e antichi esegeti di Virgilio’, ASNP 9 (1979), 539–62Google Scholar, Fowler, D., ‘The Virgil commentary of Servius’, in Martindale, C. (ed.), The Cambridge Companion to Virgil (Cambridge, 1997), 73–8CrossRefGoogle Scholar, Liveley, G., Narratology (Oxford, 2019), 100–3CrossRefGoogle Scholar, S. Poletti, ‘Eine Frage der Perspektive. Servius und Servius auctus über Vergils subjektiven Stil’, in U. Tischer, Th. Kuhn-Treichel and S. Poletti (edd.), Sicut commentatores loquuntur. Authorship and Authority in Ancient Commentaries on Poetry (Turnhout, forthcoming).

9 For ‘animo (ablative) + participle in genitive singular’, see e.g. Tib. Claud. Donat. on 2.178 reprehendentis scilicet animo faciebat […]; 3.266 […] quae tamen ipsa instruentis animo, non arguentis ingerit. For ‘animus + future participle’, see e.g. Donat. on Hec. <489> 39.1 reducturi animum; Tib. Claud. Don. on Aen. 1.131 animus audituri.

10 I thank Prof. Stok and Dr Ramires for kindly sharing with me their collations for the passages of Servius on the Bucolics that I quote here and below. Witnesses: θ (= G A), τ (= Bc Bo H Q Sc Pa Le Pc), B, γ (= E M Pb Y Z), σ (= V W); for the Bucolics there are no Δ-witnesses. I give the apparatus criticus only for the section I analyse (quomodo … caret). I rely on the sigla of Murgia (n. 1), 199–207 and Kaster and Murgia (n. 1), 1–2 and on Murgia's stemma. I add the following sigla: Bc = Oxford, Bodleian Library, Auct. T.1.25, saec. XI1; Bo = Boulogne-sur-Mer (Pas-de-Calais), Bibliothèque des Annociades, 186 (358), saec. XI; G = Glasgow, University Library, Hunterian Museum, U. 6. 8 (290), saec. IX/X; Le = Leiden, Universitaire Bibliotheken, Voss. lat. F. 25, saec. X. The editors quoted in the apparatus criticus are Guarinus Veronensis, Servii Commentarii in Vergilii Maronis opera (Venice, 1471), Cennini, P., M. Servii Honorati Commentarii in tria Virgilii opera: Bucolica, Georgica et Aeneidem (Florence, 1471–1472)Google Scholar, Stephanus, R., P. Virgilii Maronis opera (Paris, 1532)Google Scholar, Fabricius, G., Publii Vergilii Maronis opera (Basel, 1551)Google Scholar, Masvicius, P., P. Virgilii Maronis opera (Leeuwarden, 1717)Google Scholar, Burman, P., P. Virgilii Maronis opera (Amsterdam, 1746)Google Scholar, Lion, H.A., Commentarii in Virgilium Serviani (Göttingen, 1826)Google Scholar. Thilo printed sed before ratione as a conjecture by Stephanus, but this reading is already found in MSS Bo and H and in Cennini's edition.

11 diximus indicates that the topic was discussed in Servius’ note on lines 4–5 solus, which is unfortunately lost in a lacuna (Ecl. 1.37–2.10: see G. Ramires and F. Stok, ‘La lacuna del commento di Servio ad Ecl. 1.37–2.10’, RHT 12 [2017], 141–60). In the Philargyrian scholium (rec. I) to line 6 inani, there is a reference to the loneliness of Corydon: INANI idest pro ‘nihil sibi procurans contra absentem loquebatur’ (H. Hagen, Appendix Serviana [Leipzig, 1902], 33); cf. also the Bern-Scholia (H. Hagen, Scholia Bernensia [Leipzig, 1867], 754).

12 On the reference to the Epicurean doctrine and Cicero, see Setaioli, A., ‘Interpretazioni stoiche ed epicuree in Servio e la tradizione dell'esegesi filosofica del mito e dei poeti a Roma (Cornuto, Seneca, Filodemo) II’, IJCT 11 (2004), 3–46Google Scholar, at 36–7.

13 Thilo offers no indication of this, but see already Burman (n. 10), ad loc.: ‘haec nota pertinet ad vs. 60’ (similarly H. Georgii, Die antike Vergilkritik in den Bucolica und Georgica [Leipzig, 1904], 226 and Setaioli [n. 12], 36). Cf. also the parallel of Serv. on Aen. 6.465 SISTE GRADVM discedere eam datur intellegi (at line 466 Aeneas asks Dido precisely ‘quem fugis?’).

14 With ‘nunc dicit + quotation’ Servius usually gets back to the lemma, but here the quotation does not go so far as to include the lemma and works only as a cross-reference to the ratiocinatio/obiurgatio of lines 56–7.

15 It is often possible to detect the merging of different scholia in the commentary of Servius by means of comparison with the auctus. The conflation in Serv. on Ecl. 2.58 may also be due (at least partially) to a rearrangement of this section in the transmission of the commentary.

16 Cf. Georgii (n. 13), 226–7.

17 Serv. on Aen. 1.272 quomodo trecentos annos dicit, cum eam quadringentis regnasse constet sub Albanis regibus? sed cum praescriptione ait ‘tercentum’ […] (compare the sed in MSS Bo H at Serv. on Ecl. 2.58); G. 2.460. See also Serv. auctus on Aen. 4.696 harum rerum ratio sic redditur in the answer to a quomodo-question. On the quaestiones, cf. É. Thomas, Scoliastes de Virgile: Essai sur Servius et son Commentaire sur Virgile (Paris, 1880), 247–57, P.C. Burns, ‘The Vatican scholia on Virgil's Georgics’ (Diss., Toronto, 1974), 190–9, Poletti, S., ‘Due note testuali (Serv. auct. Georg. 2, 148; 434)’, Hermes 146 (2018), 373–80CrossRefGoogle Scholar, at 376–8.

18 I do not discuss the variant of word order.

19 For ‘nam ait + quotation’ in parentheses, see e.g. Serv. on Aen. 9.630 sic Homerus Thersiten a tergo uulneratum dicit usque ad praecordia (nam ait μετάφρενα) quia eum stultum induxerat. Even with the quomodo-variant, nam ait solus montibus et siluis should be printed in parentheses (see the punctuation above).

20 Possibly quod was a correction of quo(modo) in the archetype and was then added in different positions in some parts of the tradition. The reading of MS Y can hardly be a conjecture. There is a varia lectio (of no value) also in the text of the Reg. Lat. 1495: quomodo uel cum eum dicit discedere, etc. (see Thilo's [n. 1] apparatus criticus).

21 This form is also used in other contexts, e.g. to confute other commentators’ opinion: Serv. on Aen. 2.7 illud autem quod Asinius Pollio dicit caret ratione (see also 2.557, 9.410, 9.412, 12.183). For instances of ratione (non) caret without quod-clause as a subject, see Serv. on Aen. 1.642, 663; 7.457; 11.721; 12.725; and on G. 1 praef.

22 On modo = hic, hoc loco, νῦν (TLL 8.1311.61–73), typical of Servius, see Thomas (n. 17), 148, Bulhart, V., ‘Textkritisches zu Servius’, Mnemosyne 6 (1953), 64–5CrossRefGoogle Scholar, A. Uhl, Servius als Sprachlehrer (Göttingen, 1998), 541 n. 181.

23 For quod modo in Servius, see e.g. Serv. on Aen. 2.492 […] unde est quod modo dixit ‘me primam’; 10.272. Particularly interesting is the formulation of Serv. on Aen. 9.367 (Murgia's edition): VRBE LATINA non est contrarium illi loco ubi ait [7.600] ‘saepsit se tectis rerumque reliquit habenas’ quod [quod Δ Pa Pc W : quo F J1 Γ] modo a Latina urbe auxilia uenire commemorat. intellegimus enim Latinum in principio discordiae et tumultus paululum se abstinuisse. Here the quo of F J1 Γ is obviously rejected by all editors.

24 See A. Cucchiarelli, Publio Virgilio Marone, Bucoliche (Rome, 2012), ad loc. Further bibliography: Radke, G., ‘Vergils Cumaeum carmen’, Gymnasium 66 (1959), 217–46Google Scholar; A. Wlosok, ‘Cumaeum carmen (Verg., Ecl. 4, 4): Sibyllenorakel oder Hesiodgedicht?’, in A. Wlosok, Res humanae, res divinae (Heidelberg, 1990), 302–19.

25 See e.g. Radke (n. 24), 234 and 240, M. Irvine, The Making of Textual Culture (Cambridge, 2006), 152, O. Waßmuth, Sibyllinische Orakel 1–2 (Leiden and Boston, 2011), 34.

26 Daspet, F., Traduction du Commentaire de Servius aux Bucoliques de Virgile (Gradignan, 2007), 34Google Scholar (‘[les verses prophétiques] de la Sibylle, qui était de Cumes et qui divisé les siècles selon les métaux; elle a prédit aussi …’) presupposes Sibyllae quae and a different punctuation (see below).

27 This interpretation is present in Philarg., the Bern-Scholia and ps.-Prob. ad loc. and is usually rejected by modern scholars (contra, Radke [n. 24], with analysis of the scholia at 239–40).

28 See Philarg. ad loc. The existence of different Sibyllae is acknowledged by Servius in the commentary on the Aeneid: Serv. on Aen. 3.445, 6.36, 6.72, 6.98 CVMAEA SIBYLLA bene addidit propter discretionem. At Ecl. 4.4 he takes the story of the different Sibyls for granted, hence the concise explanation quae Cumana fuit.

29 Philarg. on Ecl. 4.4 rec. II quidam interpretantur Sibyllam, quae fuerit Cumana, quae praedixit futura; Bern. Schol. on Ecl. 4.4 Cymaei, Sibylliaci. Cymaei, quia Sibylla quattuor deorum descripsit regna; quae Cymaea dicitur de monte Cymo […] Alii Sibyllam, quae Cymaea fuit, intellegunt, quae quattuor saecula libris suis digessit. […] (see Hagen [n. 11 (1902)]); Lib. Glos. Sibille, quae quattuor seculorum ordinem scripsit (A. Grondeux and F. Cinato [edd.], Liber glossarum digital [Paris, 2016]: http://liber-glossarum.huma-num.fr). Cf. also the mention of the Sibyl in Serv. on Ecl. 4.10 (quoted above). For his correction quod … quae (see below) Thilo is relying on some parallels from Virgilian exegesis as well: Philarg. on Ecl. 4.4 rec. I quidam interpretantur Cymaei Sibyllam quod fuerit illa Cumaea, quae futura praedixit; see also Thilo, G., ‘Beiträge zur Kritik der Scholiasten des Vergilius’, RhM 15 (1860), 119–54Google Scholar. There is a similar formulation with quod in ps.-Prob. on Ecl. 4.4.

30 Cf. Thilo's apparatus criticus: ‘expectatur vel sibyllae vel sibyllini quod Cumanae fuit quae saecula e.q.s.’. I believe that Thilo's conjecture Sibyllae presupposes the paradosis quae Cumana, etc. (and not quod Cumanae). Sibyllae in place of Sibyllini is already found in some early editions, e.g. in J. de Mareschal, Opera Vergiliana (Lyon, 1528) (in the form Cumaei Sibyllae quae Cumana fuit. haec saecula … diuisit et dixit quis …); after Thilo's edition, in Cumont, F., ‘La fin du monde selon les mages occidentaux’, RHR 103 (1931), 29–96Google Scholar, at 44 n. 2. Geffcken, J., ‘Die Hirten auf dem Felde’, Hermes 49 (1914), 321–51Google Scholar, at 326 and Funaioli, G., Esegesi virgiliana antica (Milan, 1930)Google Scholar, 21 n. 3, 222 consider the paradosis corrupt. Funaioli suggests the parallel of Philargyrius to correct Servius’ text (see previous note).

31 Guarinus (n. 10), followed by other editions depending on him.

32 Corssen, P., ‘Die vierte Ekloge Virgils’, Philologus 81 (1925), 26–71Google Scholar, at 33. While leaving out Sibyllini, some scholars quote the scholium only with Sibylla (e.g. Kraggerud, E., ‘Further problems in Vergil’, Symbolae Osloenses 65 [1990], 63–77CrossRefGoogle Scholar) or Sibylla in parentheses (e.g. Kraus, W., ‘Vergils vierte Ekloge’, ANRW 11.31.1 [1980], 604–45Google Scholar, at 610 and Cancik, H., Gesammelte Aufsätze I [Heidelberg, 2008], 99)Google Scholar.

33 I would exclude other possible corrections of Sibyllini (e.g. Sibyllae. nam … : see sibilla [sic] nam, a correction in MS Pc). Servius uses Sibyllinus as a gloss also in Serv. on Aen. 6.72 (TVAS SORTES Sibyllina responsa). The gloss Cumaei–Sibyllini is already found in August. Ep. 258.5 quod ex Cymaeo, id est ex Sibyllino carmine se fassus est transtulisse Vergilius (Epistulae ad Romanos inchoata expositio 3), who is possibly criticizing the interpretation of Cumaeus as a reference to Hesiod: see Roessli, J.-M., ‘Augustin, les sibylles et les Oracles sibyllins’, in Fux, P.-Y., Roessli, J.-M. and Wermelinger, O. (edd.), Augustinus Afer (Fribourg, 2003), 263–86Google Scholar, at 264 n. 5.

34 If the omission after diuisit goes back to the compilation of DS, I see at least two possible explanations: 1. DS transcribed the first (corrupted) part of the scholium from S and decided to omit the rest, even if it was probably present in his sources, both S and D (the reason of this omission remains unclear); 2. dixit etiam quis, etc. is an original expansion by S of a D-gloss ending with diuisit, which DS closely reproduces. In this second scenario, it is hard to say how we should consider the corruption Sibyllini quae and at which level it may have occurred (perhaps DS copied the first part of the scholium from S and omitted the rest according to D; possibly the omission of a Sibyllae/Sibylla after Sibyllini occurred polygenetically in the traditions both of DS and of S). On the relation S-D-DS, cf. Kaster and Murgia (n. 1), xi, xx–xxviii.

35 If DS reflects an original D-gloss with a genitive Sibyllae (see previous note) and if this D-gloss was present to S, an S-text such as Sibyllini. Sibylla quae … should then presuppose some original rearrangement of the scholium by Servius. This possible relation D–DS cannot be taken for granted, but it does invite us to take into serious consideration solutions, including the genitive Sibyllae as well.

36 A conjunction (Sibylla enim; nam Sibylla), though not strictly necessary, would perhaps make the syntax smoother.