^{page 42 note 1} Cf. Gellius, , Noct. Att. xi. 2. 5.Google Scholar

^{page 42 note 2} Frag. 231–4 in Warmington's Loeb edition of Ennius, from which all the fragments of Ennius are cited throughout. The second line means the rejection of the Saturnian and not of the title vates. There is no evidence that the semi-graecus (Suetonius, Gramm. i)Google Scholar Livius Andronicus called himself a vates, while Naevius calls himself a poeta in his epitaph (Warmington, , Remains of Old Latin, Loeb, vol. ii, p. 154)Google Scholar and is described as such by Plautus (Mil. Glor. 211Google Scholar, if this reference is to Naevius) and by the Metelli in their famous gibe. Vates in the sense of poet was an Augustan neologism. (Cf. the articles mentioned by Latte, K., Römische Religionsgeschichte (Munich, 1960), 268Google Scholar: I am hoping to discuss this point further in a separate article along the lines indicated by Dahlmann.)

^{page 43 note 1} Cf. frag. 6–7 of the Satires, ‘Enni poeta salve’, etc., and Cicero, Pro Archia, 18 ad fin., on the difference of poetry from the other arts, mentioning Ennius.

^{page 43 note 2} This is to accept the traditional view that modos in this passage means ‘metres’, but surely this view is wrong, (a) The Odes are not written in ‘Italian metres’. (b) Similar claims to originality in other poets (e.g. Propertius, iii. 1. 3–4: Butler and Barber give a list of parallels in their note) are not concerned with metre as such, (c) Epistles, i. 19. 23–25Google Scholar, where the key words are Parios, numeros animosque, Archilochi, shows that here too the proper names are vital: Horace means, ‘I was the first to write Latin poetry in the spirit of Alcaeus’. That this is what Aeolium carmen signified for him he had already made clear at Odes, i.Google Scholar 32.5. There is no need to excuse Horace for failing to refer to Catullus. He thought that poetry was a question of attitudes, not of glyconics and pherecrateans. So did Ennius.

^{page 43 note 3} Frag. 229–30.

^{page 43 note 4} Cf. Skutsch in Pauly-Wissowa, s.v. Ennius, col. 2604, 18 f. If the dream motif is from Callimachus, the transformation which it underwent at the hands of Ennius is characteristic for the history of Roman poetry, as is well brought out in many passages of L'influence grecque sur la poésie latine de Catulle à Ovide (Fondation Hardt, Vandœuvres-Genève, 1953), notably in the essay on Virgil.

^{page 43 note 5} Warmington, , Remains of Old Latin, vol. i, p. 6, note (a).Google Scholar

^{page 43 note 6} Cf. Altheim, F., A History of Roman Religion (London, 1938), 300 ff.Google Scholar One need not accept all Altheim has to say in order to realize that his general insistence on the importance of Ennius' dream is correct.

^{page 44 note 1} Frag. 235. Cf. also Suidas, quoted by Skutsch in P–W. col. 2598, 54.

^{page 44 note 2} We should not be too disdainful of the Saturnian. We had a similar clash in our own literature between imported methods of scansion and the native variety, e.g. between the alliterative verse of Langland and the syllabic scansion of Chaucer, and between principles of stress and quantity. Fortunately neither side won outright victory. ‘English metre … is seen to have been brought to artistic perfection by the fusion of several modes’ (Hamer, E. in Chambers' Encyclopedia [London, 1959], vol. 9, p. 348Google Scholar, 2nd col.). The double structure of the lines of Piers Plowman has obvious affinities with the Saturnian, and the beauties of the one show what the other might have become. After all, as the caesura perhaps indicates, the hexameter must have risen from similar beginnings.

^{page 44 note 3} Although Suetonius (Gramm. i), who gives us this information, attributes the books to a younger Ennius, a grammarian, Skutsch is clearly on the right lines when he rejects the existence of this figure. His activity at the end of the second century is an invention by authors too preoccupied with the beginnings of the Golden Age. See some wise remarks by ProfFraenkel, , Horace (Oxford, 1957), 124Google Scholar and footnote: ‘Behind his (Ennius') spirited writing there lies a great effort of sober thought and selective planning.’ The scholarly compromise suggested by Bardon, H. in La littérature latine inconnue (Paris, 1952), i. 33–34Google Scholar, does not affect the substance of this point.

^{page 45 note 1} These details are from Snell, B., Griechische Metrik (Göttingen, 1957), 6–9.Google Scholar

^{page 45 note 2} 1 in 50 lines of Homer: 5 in about 550 lines of the Annales.

^{page 45 note 3} Too much should not be made in this respect of the ‘natural heaviness’ of Latin. Ennius has more dactyls than Catullus: cf. the article by Patzer, H., ‘Zum Sprachstil des neoterischen Hexameters’, Museum Helveticum (1955), 77 f.Google Scholar

^{page 45 note 4} Crusius-Rubenbauer in ch. 2 of their Römische Metrik (Munich, 1961)Google Scholar discuss pre-Ennian prosody. As for spelling we may see in the introduction by Ennius of double consonants an example of his Oscan cor at work, since they were a feature of that dialect (e.g. meddix, Ann. frag. 290) but not originally of Latin.

^{page 46 note 1} Frag. 139.

^{page 46 note 2} There is a very interesting comparison of Plautus and Ennius in their use of language in Leo, F., Geschichte der römischen Literatur (Beriin, reprint of 1958), i. 182 ff.Google Scholar Although Leo emphasizes the archaic and even non-Latin elements in Ennius, it still seems clear that his language is closer to ordinary Latin than the Greek epic dialect is to ordinary Greek.

^{page 46 note 3} This mingling of literary genres was noted by Kroll, W. in his Studien zum Verständnis der römischen Literatur (Stuttgart, 1924), 202–24Google Scholar, as one of the characteristics of Latin literature. A good example of it is Ennius' fondness for periphrastic expressions of utterance (the evidence is too long to quote here: see E. Norden's commentary on Aeneid vi (4th ed., Stuttgart, 1957)Google Scholar at lines 55, 76, 160, 628, and his Appendix I, 2. 373). Although these expressions are found in Homer (e.g. ὄπα ἱέναι in Iliad iii. 152Google Scholar, and the formula ἔπεα πτερόεντα προσηύδα) they are much more characteristic of the language of tragedy (cf. Ag. 37Google Scholar: Hipp. 418, 1074).Google Scholar I have examined this topic at greater length in an article which I hope to complete shortly.

^{page 46 note 4} Horace makes a very acute criticism here of tragic style which we may regard as equally applicable to the Annales: ‘et placuit sibi, natura sublimis et acer’ (Ep. ii. 1. 165). The latter half of the line would be a very good description of Ennius: unfortunately ‘placuit sibi’ reveals the complacency which destroys art.

^{page 47 note 1} OCD s.v. Tacitus, para. 12. Cf. Norden, , Antike Kunstprosa (Stuttgart, 1958), i. 328 and 331Google Scholar on Tacitus the tragedian. Because we read Caesar and Cicero most often, and because of the loss of so much Latin prose, we think of the archaizing and poeticizing nature of Latin prose outside these two authors as something abnormal. The fact is of course that it is precisely because they were able to resist the tide that they deserve our admiration. But they were not typical.

^{page 47 note 2} Cf. Buchner, K., Romische Literaturgeschichte (Stuttgart, 1957), 85 and 142.Google Scholar

What is said on these two pages must be taken as part of one process.

^{page 47 note 3} The quotation and examples are from Norden, E., Die römische Literatur (Leipzig, 1954), 17.Google Scholar

^{page 47 note 4} Cf. Axelson, B., Unpoetische Wörter (Lund, 1945), ch. 4.Google Scholar

^{page 48 note 1} Skutsch, , loc. cit., col. 2619, 33.Google Scholar After coming to the conclusion that this great scholar was wrong here, I was encouraged to read the footnote on p. 195 of Leo's discussion of Ennius (cited above, p. 46, n. 2). In this note Leo suggests that ‘between Plautus and Terence a reform of technique (i.e. in the metre of dialogue) involving abolition of hiatus must have taken place … one ought perhaps to think of Ennius …’. If Ennius reformed dialogue, why not lyric?

^{page 48 note 2} Loc. cit., line 56.

^{page 48 note 3} Frag. 38–49.

^{page 48 note 4} Op. cit., para. 173.

^{page 48 note 5} No doubt this statement requires qualification: it excludes for example sim/arce in the famous cretics from the Andromache (frag. 94–100)—where a question mark intervenes. No account is taken either of hiatus within the colon: e.g. in this same passage it expresses the hesitant, distraught utterance of the bewildered mother with marvellous force. The presence of synapheia is also revealed if we rearrange Cassandra's lines at 73–75 to form an iambic senarius followed by anapaestic dimeters—‘eheu videte iudicavit inclitum|iudicium inter deas tres aliquis|quo iudicio Lacedaemonia|mulier furiarum una adveniet.’ If iudicavit has its classical scansion rather than a long final syllable we have here another example of purity of structure in Ennius' senarii. Deas is probably an instance of synizesis rather than brevis brevians: if it were the latter it would be unique in Ennius' lyrics (‘adest adest fax’, so often cited, is not strictly speaking lyric: contrast Crusius, , op. cit., para. 108Google Scholar, where we learn that brevis brevians is at its commonest in Plautus' anapaests). In line two of the anapaests Lacedaemonia might have a short final syllable (like Pelopis in Accius, frag. 292) but the poet who allowed aquila to retain a long final syllable in Ann. 151 probably did not scruple here to let Lacedaemonia keep its Greek quantities. (The pregnant use of aliquis is perhaps another Greek touch.) Cf. Jebb on Antigone 751.Google Scholar

^{page 48 note 6} See Accius, frag. 277–9 and Warmington's Loeb note. The passage is also mentioned by Crusius, , op. cit. 88, top.Google Scholar

^{page 49 note 1} Cf. Fraenkel, E., Elementi Plautini in Plauto (Florence, 1960), 324.Google Scholar

^{page 49 note 2} The analyses of Plautine cantica given by Crusius in paras. 172 ff. are very striking. Horace's criticisms (A.P. 270 ff.)Google Scholar, the result of the Ennian revolution, really bring the absurdity of the situation home—numerosus Horatius unable to comprehend the numeri innumeri.

^{page 49 note 3} Val. Max. iii. 7. II.

^{page 49 note 4} The self-satisfied tone of the Pragmatica (Warmington, Loeb, , p. 588)Google Scholar rather gives the game away.

^{page 49 note 5} For all the undoubted power of the tragic fragments of Ennius we should not overlook the feeling which makes Skutsch (col. 2596, 67) speak of the ‘coldly rational’ element in them. At 282–3 Warmington bids us note the ‘thoroughly Roman’ device of rhetoric introduced by Ennius. We must alter ‘Roman’ to ‘Hellenistic’ and note that rhetoric killed tragedy. (See some good remarks by Beare in OCD, s.v. Drama, Roman, para. 6. Leo remarks (op. cit. 195): ‘Ennius' tragic diction is based on turning style into rhetoric.’ He has some good examples on p. 191.) The comparative poverty of metres in Terence and, so far as may be judged, in Pacuvius and Accius, is also worthy of attention. Tragedy at least seems to have paid for living beyond its means by having only anapaests on the menu most of the time (the metre in which ‘contradiction between accent and ictus was most frequent’—Crusius, para. 107).Google Scholar