^{page 126 note 1} Those of book 6 in sections 7–14 run 1138 ff., 1145 ff., 1182 ff., 1226 ff., 1178 ff., 1219ff. The parallel citations of the third Georgic, on the other hand, are quite random: 478 ff., 505 ff., 503 and 500 ff., 509 ff., 548 ff., 546 ff.

^{page 127 note 1} ‘De Aeschyli tragoediis fata Aiacis et Teucri complexis’, Opuscula, vii. 365 ff.Google Scholar

^{page 127 note 1} Cf. Welcker, F. G., Die griechischen Tragödien mit Rücksicht auf den epischen Cyclus I (Rh. M. Suppl. ii. 1 [1839]), p. 39Google Scholar; Ribbeck, O., In tragicos Romanorum poetas coniectanea—specimen I (Diss. Berlin, 1849), p. 19Google Scholar; Die römische Tragödie im Zeitalter der Republik (Leipzig, 1875), pp. 368Google Scholar ff.; Mueller, L., De Accii fabulis disputatio (Berlin, 1890), pp. 19Google Scholar ff; Leo, F., De tragoedia Romana obseruationes criticae (Progr. Göttingen, 1910), p. 12Google Scholar (= Ausgewählte kleine Schriften, i. 201Google Scholar); Geschichte der römischen Literatur, p. 398Google Scholar and n. 1; Warmington, E. H., Remains of Old Latin, ii (London, 1936), pp. 358Google Scholar ff; Klotz, A., Scaenicorum Romanorum fragmenta, i (Munich, 1953), pp. 214 ff.Google Scholar

^{page 127 note 3} Subsequent opinion has differed widely on Accius' immediate model.

^{page 128 note 1} Tusc. 4. 23Google Scholar. 52 ( = inc. inc. trag. 64–66Google Scholar). Cf. Sophocles, , Aias 466 ff.Google Scholar

^{page 128 note 2} At Off. 3. 26. 98 Cicero quotes a speech from a debate between Ulysses and Ajax but there is no means of determining whether the play was Accius' Armorum indicium or the homonymous play of Pacuvius.

^{page 128 note 3} Euripides restituius, i (Hamburg, 1843), p. 196.Google Scholar

^{page 128 note 4} 4. 29. 40 (= 1066 N.²).

^{page 128 note 5} Ribbeck, O., Die römische Tragödie, p. 344Google Scholar; Mueller, L., De Accii fabulis, p. 61Google Scholar; Leo, F., De tragoedia Romana, p. 10Google Scholar f. (= Ausgewählte kleine Schriften, i. 199Google Scholar f.); Buchwald, W., Studien zur Chronologic der attischen Tragödie 455 bis 431 (Diss. Königsberg, 1939), pp. 17 ff.Google Scholar

^{page 128 note 6} See 697, 698 N.².

^{page 128 note 7} Handley, E. W. and Rea, John, The Telephus of Euripides (University of London Institute of Classical Studies, Bulletin Suppl. V [1957]), Part II PP. 26, 29–30, and 39.Google Scholar

^{page 129 note 1} 696 N.². Handley, , op. cit., p. 29Google Scholar, suggests that Accius changed die order of events in Euripides' prologue, as did Ennius in the case of the Medeia. But Ennius here preserves a stylistic link with his original: utinam ne … accedisset= The only other early Latin translations whose beginnings are preserved proceed similarly: Livius Odyssia 1 uirvm mihi Camena insece uersutum =Accius, , Phoenissae 581 ff. sol qui … explicas, quianam…=Google Scholar

^{page 129 note 2} For further confirmation of the common view see Timpanaro, S., ‘Per una nuova edizione di Ennio’, S.I.F.C. xxii (1947), 74.Google Scholar

^{page 129 note 3} Welcker, F. G., Die griechischen Tragödien III (Rh. M. Suppl. ii. 3 [1841]), p. 1394Google Scholar; Ribbeck, O., Die römische Tragödie, p. 48Google Scholar, al.; Leo, F., Plautinische Forschungen, ed. 2 (Berlin, 1912), p. 90Google Scholar n. 1; Geschichte der römischen Literatur, pp. 89Google Scholar, 189, 398; Fraenkel, E., R.-E. Suppl. vi (1935)Google Scholar, s.v. Naevius, 633.Google Scholar

^{page 129 note 4} Tremulus in 7 and tempestiuus in 16 look as if they belong with gradiuus, mulciber, liquidus, tristis, and auritus in 5. 2-–. The only nouns are agmen in 4 and umbraculum in 8.

^{page 130 note 1} H. Nettleship, ‘On some of the early criticisms of Virgil's poetry’, p. lii, tables the series as: (i) additus agmen crepito horrere tremulus umbraculum (transmitto); (2) defluo discludo deductus proiectus tempestiuus; (3) (lychnus) aethra daedalus reboo; (4) camurus Mulciber petulcus (liquidus tristis auritus); (5) turicremus ueliuolus uitisator; (6) arcitenens siluicola; (7) noctiuagus nubigena. Nettleship omits uomere, pausa, machaera, asotia, malace, urus, gradiuus, and overruns Macrobius' category divisions.

^{page 130 note 2} e.g. 5. 10: …A. 10. 524, 10. 723, n. 191, 11. 486, *9. 435*; 5. 11: *A. 1. 430*, 1. 198, 2. 626, 3. 513, 4. 365…

^{page 130 note 3} 226. 4 (Lindsay1). Strzelecki, L., Quaestiones Verrianae (Warsaw, 1932), p. 81, argues that these pieces come from Verrius' own excerpting of ancient texts.Google Scholar

^{page 130 note 4} 59. 26 (Lindsay¹).

^{page 130 note 5} See J. Vahlen, Ennianae poesis reliquiae ², p. Ixvii; Rychlewska, L., ‘De Verriana hexametros afferendi ratione’, Eos xliii (1948/1949)Google Scholar, fasc. 1, 186 ff.Google Scholar

^{page 130 note 6} 1. 196. 7 (Hertz).

^{page 131 note 1} See Jeep, L., ‘Priscianus; Beiträge zur Überlieferungsgeschichte der römischen Literatur, II’, Philologus Ixviii (1909), 1–51.CrossRefGoogle Scholar

^{page 131 note 2} Cf. Priscian, , Inst, gramm. 1. 212. 21. There are also coincidences between material in the fifth and sixth books of the Institutiones and material in Festus and Paulus articles which cannot be accidental; e.g. 1. 204. 13 = Festus 166. 11; 1. 231. 13 = Festus 290. 35.Google Scholar

^{page 131 note 3} Contrast the form of quotation at 3. 19. 5 (Naeuius poeta in Bella Punico), where Verrius is quite certainly the ultimate source. On the significance of the various methods of quoting Naevius' epic see Strzelecki, L., De Naeuiano Belli Punici carmine quaestiones selectae, pp. 1 ff.Google Scholar

^{page 131 note 4} 6. 4. 7 Ennius in Melanippe … Lucretius in sexto …; 6. 4. 18 Ennius in nono… Lucretius in quinto … Lucilius in prima…; 6. 5. 14 Pacuuius in Paulo … Accius in Philoctete … idem in Minotauro…

^{page 131 note 5} On Festus' order see Strzelecki, L., Quaestiones Verrianae, pp. 81 ff.Google Scholar

^{page 131 note 6} Lindsay, W. M., Nonius Marcellus' Dictionary, p. 2, makes a medieval editor responsible for this.Google Scholar

^{page 131 note 7} Nettleship, H., ‘On some of the early criticisms of Virgil's poetry’, pp. 1Google Scholar ff., was wrong to see any significance in the fact that Nonius' articles on crepitare, deducere, proicere, boere, camerus, and trisiis are illustrated by the same passages of Virgil as occur in Satunalia 6. 4–5Google Scholar. The passage illustrating camerus at 30. 8 comes in a series of three from the Georgics and must be considered to derive from Nonius' own excerption of Virgil's poems (see Lindsay, W. M., op. cit., p. 13Google Scholar). While the other passages do not readily reveal their origins, there is some evidence that Nonius was able to quote Virgil's poems from memory (see W. M. Lindsay, ibid., p. 104).

^{page 131 note 8} See W.M. Lindsay, ibid., p. 101.

^{page 132 note 1} See Strzelecki, L., Quaestiones Verrianae, PP. 3–23.Google Scholar

^{page 132 note 2} Poetae scenici Latinorum, v (Leipzig, 1834), p. 96.Google Scholar

^{page 132 note 3} Q.. Enni carmimtm reliquiae: accedunt Cn. Naeui Belli Poenici quae supersunt (St. Petersburg, 1884), p. 167.

^{page 132 note 4} Der saturnische Vers (Berlin, 1905), p. 61 n.Google Scholar

^{page 132 note 5} Naevius poeta (Catania, 1945): see ed. 2 (Florence, 1950), p. 201.Google Scholar

^{page 132 note 6} Leipzig, 1852, p. 11.

^{page 132 note 7} Gnomon iv (1928), 689Google Scholar, reviewing Morel, W., Fragmenta poetarum Latinorum epicorum et lyricorum praeter Ennium et Lucilium (Leipzig, 1927).Google Scholar

^{page 132 note 8} II Bellum Poenicum e l'arte di Nevio (Rome, 1955), pp. 93 ff. Nevertheless Mariotti does not go so far as to print it among his fragments of the Bellum Punicum.Google Scholar

^{page 132 note 9} Cn. Naeuii Belli Punici carminis quae supersunt, p. 100Google Scholar. See also ‘Miscellanea Naeuiana’, Eos xlix (1957/1958), 65.Google Scholar

^{page 132 note 10} (1) 6. 8. 3: Vergilius alio in loco (= A. 5. 372) … et item alibi (= A. 5. 401). Cf. 5. 7. 9 (ψ … 164); (2) 6. 4. 14: (= A. 3. 699) … ut alibi ait (= A. 10. 588). Cf. 5. 12. 11 (π851 … X 364); (3) 5. 21. 1: (= G. 4. 380) … et alibi (= A. 5. 77). Cf. 5. 3. 9 (∈ 297 … Θ 330). Strzelecki's argument (Eos xlix [1957/1958], 65) that alibi at 6. 5. 8 could not mean ‘in another work by Naevius’ goes too far.Google Scholar

^{page 133 note 1} Cf. Ennius, , Ann. 129Google Scholar hic occasus datust at Oratius inclutus saltu; 136 quande tuas omnes legiones ac popularis; 355 quippe Solent reges omnes in rebus secundis; 422 qui clamos oppugnantis uagore uolanti, all quoted by Festus. Of the same character is 173 quod per amoenam urbem leni fluit agmine flumen, quoted at Saturnalia 6. 4. 4 (unless quod be taken with flumen).Google Scholar

^{page 133 note 2} Cf. Bassus, Caesius, G.L. vi. 265. 10 K.: nostri autem antiqui, ut uere dicam quod apparet, usi sunt eo non obseruata lege nee uno genere custodito, ut inter se consentiant uersus, sed praeterquam quod durissimos fecerunt, etiam alios breuiores, alios longiores inseruerunt, ut uix inuenerim apud Naeuium, quos pro exemplo ponerem.Google Scholar

^{page 133 note 3} So, apparently, Delrio. Aulus Gellius' method of introducing one of the pieces— ex libro qui Scipio inscribitur (4. 7. 3)—seems to have counted against this theory. But for liber = ‘play’ cf. Propertius 3. 21. 28, Persius 1. 76, Quintilian Inst. 1. 10. 8.

^{page 133 note 4} Skutsch, F., Ennius, 2599.Google Scholar

^{page 133 note 5} This seems to have been the common view when Vahlen published his first edition of Ennius in 1854. It was maintained by Mueller, L., Quintus Ennius, pp. 107Google Scholar ff., and rejected by Vahlen in his second edition (cf. p. ccxvi). Corte, F. Delia, ‘Intorno alle saturae di Ennio’, Atti Ace. di Torino lxxi (1935/1936), t. II, 198 ff., resurrected it.Google Scholar

^{page 133 note 6} The entry in the Suda lexicon which Vahlen prints as fragment I could, it seems to me, refer either to a passage of the Annales or to the prologue of a fabula praetexta. Fragments II–V ought to be assigned to the Annales.

^{page 133 note 7} The promise at 6. 5. 1 to illustrate the archaic use of gradiuus is not fulfilled, at least not in our text of the Saturnalia.

^{page 133 note 8} Supra, p. 131.Google Scholar

^{page 134 note 1} Cf. Ann. 285 densantur campis horrentia tela uirorum; 393 horrescit telis exercitus asper utrimque. There are several verses in the Annates lacking a normal middle caesura: 43, 48, 122, 123, 230, 522.Google Scholar

^{page 134 note 2} Cf. Ann. 557 interea fugit albus iubat Hyperionis cursum; 558 inde patefecit radiis rota Candida caelum; Virgil A. 4. 584–5 et iam prima nouo spargebat lumine terras | Tithoni croceurr, linquens Aurora cubile.Google Scholar

^{page 134 note 3} Tr. 1010–14, Ph. 839–42.

^{page 134 note 4} Tr. 595–602, Hel. 164–5.Google Scholar

^{page 134 note 5} Medea 110–15.Google Scholar

^{page 134 note 6} 1. 196. 16; in H there is a lacuna between ortam and fuisse; R omits eam, L puts it after an, the rest after humano.

^{page 134 note 7} Cf. Aeschylus, Pr. 588Google Scholar

^{page 134 note 8} Poetae scenici Latinorum, p. 211.Google Scholar

^{page 134 note 9} Scenicae Romanorum poesis fragmenta, i (Leipzig, 1852), p. 302; similarly in the two succeeding editions of this work.Google Scholar

^{page 134 note 10} So, according to Hertz's apparatus, Dresd. Krehlii. D has Accius in I.oe.

^{page 134 note 11} Cf. supra, p. 130.Google Scholar

^{page 134 note 12} Cf. the way in which the Naples manuscript of Charisius, 79. 22 B., corrupts adsiduum Ioni to adsiduum Minoi.

^{page 135 note 1} De Vergilio, pp. 5, 8, 38.Google Scholar

^{page 135 note 2} With Saturnalia 5. 3. 1—et si uultis me et ipsos proferre uersus ad uerbum paene translates— should be compared Servius ad A. 11. 483— haec autem omnis oratio uerbum ad uerbum de Homero translala est (= Saturnalia 5. 3Google Scholar. 10); ad A. 12. 84—Homeri uersus, uerbum ad uerbum; ad A. 12. 206—Homeri locus, uerbum ad uerbum (- Saturnalia 5. 3Google Scholar. 13). With Saturnalia 5. 13. 40—sed haec et alia ignoscenda Vergilio, qui studii circa Homerum nimietate excedit modum—should be compared the remark extracted by the Verona scholiast from the commentary of Aemilius Asper ad A. 10. 559–in quibusdam dum nimio studio Vergilius ad Homerum trahitur, neque temporis, neque loci habet curam.Google Scholar

^{page 136 note 1} At Saturnalia 6. 4. 2 Lucilius is cited but not Plautus. For reasons to be discussed later Macrobius does not cite Plautus any where in 6. 4–5.Google Scholar

^{page 136 note 2} Verrius Flaccus is named by Servius Danielis at B. 7. 53 and A. 11. 143. For Servius, Verrius Flaccus, and Festus cf. Strzelecki, L., Quaestiones Verrianae, pp. 73 ff.Google Scholar

^{page 136 note 3} Festus worked somewhere between the time of Martial and that of Porphyrio (see Helm, R., R.-E. xxi [1952]Google Scholar, s.v. Pompeius Festus, 2316).Google Scholar

^{page 136 note 4} Caper is named by Servius at A. 6. 545 and A. 9. 706.

^{page 136 note 5} Cf. Servius, D. ad A 2Google Scholar. 355 = Festus 364. 1; Servius, D. ad. A. 12Google Scholar. 657, G. 4. 188 = Paulus 131. 9; Schol. Ver. ad B. 5. 88 = Festus 292. 7.

^{page 136 note 6} There is some dispute about the date of Aemilius Asper; see Thomsin, A., Étude sur le commentaire virgilien d'Aemilius Asper (Paris, 1952), pp. 17 ff.Google Scholar

^{page 136 note 7} Servius ad B. 10. 46; ad G. 2. 42; Schol. Bern, ad G. 2. 94; Servius ad G. 2. 404; ad A. 2. 241; ad A. 2. 274; ad A. 5. 591; ad A. 6. 219; ad. A. 6. 625; ad A. 6. 845; ad A. 7. 622; ad A. 9. 225; ad A. 10. 396; ad A. 11. 27; ad A. 11. 601; ad A. 12. 115.

^{page 136 note 8} Servius Danielis ad G. 1. 375; Servius ad G. 3. 293; Servius Danielis ad A. 1. 198; Servius ad A. 2. 486; ad A. 8. 631; ad A. 10. 104.

^{page 137 note 1} Servius ad A. 4. 404; A. 9. 163; A. 9. 501; A. 10. 807; A. 12. 552. Cf. Macrobius, Saturnalia 6. 1. 7 quos ab aliis traxit uel ex dimidio sui versus.Google Scholar

^{page 137 note 2} Servius ad A. 11. 608. Cf. Macrobius, Saturnalia 6. 2. 7 lineamenta paene omnia tracta sunt de descriptione pestilentiae.Google Scholar

^{page 137 note 3} Cf. Macrobius, Saturnalia 6Google Scholar. 1. 15 = Servius D. ad A. 9. 420 Ennius in prima …; 6Google Scholar. 1. 17 = Servius ad A. 12. 552 hemistichium est Ennianum; 6Google Scholar. 1. 18 = Servius D. ad A. 9. 526 est autem Ennianum …; 6Google Scholar. 1. 25 = Servius D. ad A. 9. 457 Lucretianum est; 6. 2. 7 = Servius ad G. 3. 478 hanc autem pestilentiam ordine quo diximus plenissime est Lucretius exsecutus1.Google Scholar

^{page 138 note 1} Cf. praef. 2: siquando usus uenerit out historiae quae in librorum strue CLAM TYLGO EST out…

^{page 138 note 2} They are all mentioned, for example, by Quintilian.

^{page 138 note 3} Varius was known only as a tragedian and one of the editors of the Aeneid. His Thyestes seems to have survived in at least one manuscript to the eighth century; cf. Housman, A. E., ‘The Thyestes of Varius’, C.Q. xi (1917), 42.CrossRefGoogle Scholar

^{page 138 note 4} Seeing that Verrius Flaccus' lexicon was one of the ultimate sources of 6. 4–5, one would have expected Plautus at least to appear in these chapters.

^{page 138 note 5} The monograph composed by Q. Octavius Avitus may have contained an account of Virgil's borrowings from Greek as well as from Latin authors but its reported size of eight uolumina remains strikingly large.

^{page 138 note 6} ad G. 2. 43.

^{page 138 note 7} ad G. 2. 43, A. 6. 625. Nothing like Virgil's verse occurs in our text of Lucretius. Lachmann therefore assumed a lacuna after 6. 839. Others (see Regel, G., De Vergilio, pp. 81Google Scholar f.) have suggested that Servius' Lucretii is an error for Lucilii. This is a common confusion made most unmistakably by the Laurentian manuscript of Varro's De lingua latina at 5. 17.Google Scholar

^{page 139 note 1} Saturnalia 6Google Scholar. 1, 9 gives three models of A. 4. 48256. 1. 22 three of A. 8. 596; 6. 1. 44 two of A. 1. 691. Cf. Servius, ad A. 4. 404; A. 10. 396; A. 11. 601.Google Scholar

^{page 139 note 2} Willis, J. A., ‘De codicibus aliquot manuscriptis Macrobii Saturnalia continenibus’, Rh. M. c (1957), 162, writes culmina for perdita.Google Scholar

^{page 139 note 3} ad G. 1. 375: SVSPICIENS PATVLIS locus de Varrone est; ille enim sic … The manuscripts of the so-called-Breuis expositio (ad G. 1. 397) attribute to Varro in Epimenide | Epimedine (Ephemeride Bergk) some words which are similar to Aratus 938 f. in the same way as the verses of the Danieline citation are similar to Aratus 942–4, 954–7.Google Scholar

^{page 140 note 1} With turn liceat… cernere … certare | aut… circumuolitauit … decerpsit contrast Virgil, , G. 1. 374 ff. aut…fugēre … captauit … circumuolitauit … cecinere … extulit … increpuit and Aratus 943 ff.Google Scholar

^{page 140 note 2} Regel, G., De Vergilio, pp. 83Google Scholar f., argued for a lacuna after the word hirundo on the grounds that aut was an awkward connective as the text stands and that, while the first four verses of the fragment corresponded with Aratus 942–4, the second three corresponded not with 945 ff. but with 954–7. It seems rash, however, to suppose that Varro's work was an exact literal translation of the Diosemeiai; cf. Leo, F., ‘Vergil und die Ciris’, Hermes xxxvii (1902), 52Google Scholar (= Ausgewählte kleine Schriften, ii. 67).Google Scholar

^{page 140 note 3} Saturnalia. 2. 1Google Scholar; cf. C.Q. N.s. xiv(1964), 289Google Scholar. Macrobius' words at 6.1. 7—quos ab aliis traxit paene solidos—seem to reflect a slightly more careful description of the parallelisms.

^{page 140 note 4} Ann. 370Google Scholar; adduced with A. 6. 846 at Saturnalia 6. 1. 23; it is, however, possible, that Macrobius has corrupted A. 6. 846 to make it more like the Ennian verse, for the earliest codices of the Aeneid write unus qui nobis cunctando restituis rem.Google Scholar At Saturnalia 6. 1. 39 Varius De mork fr. 1— VENDIDIT HIC Latium populis agrosque Quiritum | eripuit: FIXIT LEGES PRETIO ATQVB REPIXIT is paralleled with A. 6. 621–2 VEXDIDIT HIO auro patriam dominumque potentem | imposuit: FIXIT LEOES PRETIO ATQTE REF1XIT.Google Scholar

^{page 140 note 5} In the four cases where the text and attribution of an archaic piece are transmitted jointly by Saturnalia 6. 1–2 and a scholium (6. 1. 15 Ennius, Ann. 99Google Scholar = Servius D. ad A. 9. 420; 6. 1. 17 Ennius, Ann. 161Google Scholar = Servius ad A. 12. 552; 6. 1. 18 Ennius, Ann. 174Google Scholar = Servius D. ad A. 9. 526; 6. 1. 23 Ennius, Ann. 370 = Servius ad A. 6. 845), the scholiast presents much less than Macrobius.Google Scholar

^{page 141 note 1} Servius ad G. 2. 42; schol. Bern, ad G. 2. 94; Servius ad A. 6. 219; ad A. 6. 625; ad A. 6. 845; ad A. 7. 622; ad A. 9. 225; ad A. 10. 396; ad A 12. 115.

^{page 141 note 2} Servius ad A. 6. 219 (Ennius, , Annales 155Google Scholar, and Virgil A. 6. 219); ad A. 6. 845 (Ennius, , Annales 370, and Virgil A. 6. 846).Google Scholar

^{page 142 note 1} Fragmenta poetarum Romanorum (Leipzig, 1886), p. 333.Google Scholar

^{page 142 note 2} Fragmenta poetarum Latinorum (Leipzig, 1927), p. 95.Google Scholar

^{page 142 note 3} Cf. Regel, G., De Vergilio, p. 85.Google Scholar

^{page 142 note 4} Ennianae poesis reliquiae, ed. 2, pp. cvi, 3.

^{page 142 note 5} Saturnalia 6. 1Google Scholar. 17 has Ann, . 161 Romani scalis summa nituntur opum uiGoogle Scholar as the primary quotation, Ann, . 411–12Google Scholar reges per regnum statuasque sepulchraque quaerunt | aedificant nomen summa nituntur opum ui as the secondary. Vahlen prints Servius ad A. 12. 552 as a testimonium to Ann, . 161.Google Scholar

^{page 142 note 6} Virgil, , A. 2. 237–8 scandit fatalis machina mums | feta armis = Ennius, Sc. 76–77 nam maxima saltu superauit grauidus armatis equus, qui suo partu ardua perdat Pergama. Cf. the parody at G. 3. 139–41 exactis grauidae cum mensibus errant, | non illas grauibus quisquam iuga ducere plaustris, | non saltu superare uiam sit passus.Google Scholar

^{page 142 note 7} The similarity of A. 2. 281 ff. to Ennius, , Sc. 72–75Google Scholar, is noted at Saturnalia 6. 2. 18.Google Scholar

^{page 142 note 8} At Sc. 92. So also Ribbeck at Trag. 81.

^{page 143 note 1} Regel, G., De Vergilio, p. 59Google Scholar, argues this briefly in relation to Ann, . 7.Google Scholar

^{page 143 note 2} Cf. C.Q. N.s. xiv (1964), 291.Google Scholar

^{page 143 note 3} Servius D. ad A. 1. 170; ad A. 1. 198; ad A. 1. 273; schol. Veron. ad A. 2. 687 (to be supplemented from Probus ad B. 6. 31); Servius D. ad A. 2. 797; ad A. 3. 10; ad A. 4. 9; schol. Paris, lat. 7930 ad A. 7. 123; Servius D. ad A. 9. 712.

^{page 143 note 4} Servius Danielis on A. 1. 170, A. 1. 273, A. 4. 9, A. 9. 712, the Verona scholiast on A. 2. 687, and the Paris scholiast on A. 7. 123 make remarks about Naevius' version of the myth similar to those which Varro makes at L.L. 5. 43Google Scholar and L.L. 5. 53Google Scholar and Lactantius repeats from Varro at Diu. inst. 1. 6Google Scholar. 9, and thus might depend on antiquarian literature. Nevertheless it is not out of the question that at one time, before the abbreviators got to work, these scholia resembled rather the present Danieline scholia on A. 2. 797 and A. 3. 10. Strzelecki, L., Cn. Naeuii Belli Punici carminis quae supersunt, p. 39, argues that the reference to Naevius and Ennius in the scholium on A. 1. 273 comes along with the rest of the scholium from the same antiquarian source as Festus 326. 28–330. 18; but makes the other scholia derive from a writer on Virgil's borrowings from earlier literature.Google Scholar

^{page 143 note 5} Saturnalia 6. 2. 30 sunt alii loci plurimorum uersuum quos Man in opus suum cum paucorum immutatione uerborum a ueteribus transtulit… 6. 2. 31 hie locus totus sumptus a Naeuio est ex prima libro belli Punici = Servius D. ad A. 1. 198 et totus hie locus de Naeuio belli Punici libro translatus est … ad A. 2. 797 sane adamat poeta ea quae legit diuerso modo proferre … ad A. 3. 10 amat poeta ea quae legit immutata aliqua parte … proferre.Google Scholar

^{page 143 note 6} De Vergilio, p. 70.Google Scholar

^{page 143 note 7} 6. 2. 30 et quia longum est numerosos uersus ex utroque transcribere, libros ueteres notabo, ut qui uolet illic legendo aequalitatem locorum conferendo miretur.

^{page 144 note 1} Fragmenta poetarum Romanorum, p. 45.Google Scholar

^{page 144 note 2} This is actually printed by the Harvard editors of Servius. Cf. the remarks of Fraenkel, E. (J.R.S. xxxviii [1948], 139).Google Scholar

^{page 144 note 3} Cf. Servius ad G. 3. 293 MOLLI OLITO … hie autem locus lotus de Lucretio translates est (= Saturnalia 1. 2. 3 Lucretius in prima …).Google Scholar

^{page 144 note 4} I am grateful to Professor C. O. Brink for help in the preparation of this paper.