3 rhetorician at his command, but he also has the benefit of all that has preceded him in Roman literature, as well as in that of Greece, and makes good use of it. No Latin author probably has borrowed as freely and extensively from his contemporaries and immediate predecessors. But Ovid does not lack the genuine poetic power of coining new words to meet his necessities. Something like half a thousand of these we probably owe to his invention. No noteworthy syntactical peculiarities worry the student of Ovid. His style is perfectly transparent, and as a rule the thought of each distich is complete in itself. But his even regularity was fatal to the life of elegy. "Tibullus had written naturally and feelingly on love, old age, and the country. But themes which had been by him treated simply soon became fixed conventions. Ovid, despite his clearness, contributed to the progress of artificiality. The loss of the true Tibullian simplicity in theme and the loss of the true Ovidian ease in movement are evident many generations before the elegies, at once sensuous and frigid, which were written by Maximianus in the sixth century.' 41. This is not the place to discuss the Mss. and editions of all of Ovid's works in detail. The Mss. of the Amores and Heroides. are somewhat fragmentary. The Codex Parisinus (Puteanus) 8242 (P), of the eleventh (or ninth ?) century, contains most of the Amores and the larger part of the Heroides. The Parisinus Regius 7311 (R), of the tenth century, has, besides several others of the erotic works, Amor. 1, 1, 3-2, 49. The Sangallensis 864 (S), of the eleventh century, contains the Amores as far as 3, 9, 10, with the omission of 1, 6, 46-8, 74. The Guelferbytanus (G), of the twelfth century, much corrected by a later hand (thirteenth 1 Cf. Zielinski in Philologus, Vol. 64 (1905), p. 16. 2 Cf. E. K. Rand in TAPA., Vol. 35 (1904), pp. 143 sqq.; Gansemüller in Philologus, Vol. 70 (1911), pp. 274-311 and 397-437. 3 Linse, De P. Ovidio Nasone, Vocabulorum Inventore, allows him 487; Schütte, in BPW., Vol. 12 (1892), Sp. 12, thinks the number may be increased to 514. 4 Duff, p. 611. century) contains the Heroides. An Eton fragment (E) of the eleventh century contains the Heroides up to 7, 157 only. Other excerpts or fragments may be passed over at this time except the Schedae Vindobonenses (V), beginning at 10, 14. For the Tristia and Ex Ponto, the chief Mss. besides the corrupt Laurentianus (L), eleventh century, are the Guelferbytanus (G), thirteenth century, Holkhamicus (H), thirteenth century, Palatinus (P), fifteenth century, and Vaticanus (V), thirteenth century, besides a lost Marcianus Politiani (A).1 The chief text editions of all of Ovid's works are those of Riese (2d ed., 1889 sqq.), Ehwald-Merkel (4th ed., 1888 sqq.), and Postgate's Corpus Poetarum Latinorum. The Amores have been edited in German with valuable introduction, commentary, and appendices (including useful bibliography) by P. Brandt. The editions of Palmer (1898) and Sedlmayer (1886) are most important for the Heroides. Thirteen Heroides are in the convenient English edition of Shuckburgh, with introduction and commentary. For the Tristia Owen's edition (1889) is valuable. The Epistles Ex Ponto are in a critical edition by Korn 2 (1868). Ovid's works have metrical English versions by Dryden and other poets. THE ELEGIAC DISTICH 42. The laws governing the relatively simple metrical form composed of a single dactylic hexameter followed by a single dactylic pentameter-so-called-are but few; and at first sight it would seem as if there were only a narrow margin for the exercise of originality in treatment. In the hexameter there are certain positions between which the writer must choose for his verse caesura; he is expected to employ a fair proportion of dactyls, one being regularly found in the fifth foot; the verse 1 Cf. Postgate's Corpus Poetarum Latinorum, Ehwald's Praefatio, Shuckburgh's Introduction; Owen's edition. 2 Cf. BPW., Vol. 16 (1896), Sp. 1163 sqq. should end preferably with a word of either two or three syllables; harsh elisions should be avoided. In the pentameter the end of a word should always coincide with the end of the first half of the verse; the last half of the verse must always consist of two dactyls followed by a single syllable; elisions should be sparingly employed, and at any rate harsh ones avoided. But besides such few simple principles for the government of the meter, we find that in practice there grew up various other rules, and many refinements came into vogue, so that we can trace a very interesting progress in the mode of the verse from Catullus to Ovid and can see many indications of individuality in its treatment by the various authors. The subject is too large to be discussed exhaustively here; but the student may be referred to a large body of studies, which is constantly growing, with reference to it, and encouraged to pursue his own investigations along this line. The growth of new conventional usages in this verse is seen especially in the endings of the hexameter and of the pentameter, the treatment of the verse caesura, the relative proportion of dactyls and spondees and their arrangement, in care in avoiding harsh elisions, especially those of a long vowel before a short one, middle and end rime in both hexameter and pentameter, in alliteration, repeated sounds and syllables, and other euphonic embellishments, and in the tendency, culminating in Ovid, to make each distich a complete thought in itself. Some of the results of studies along some of these various lines are given below, virtually in the form in which they were published in PAPA., Vol. 34 (1903), pp. xxviii-xxx. 1 Cf. the exhaustive studies in Hosius, p. 180. Ovid avoids eliding monosyllables almost entirely; cf. Winboldt, Latin Hexameter Verse, p. 177. 2 Cf. the richly illustrated article of B. O. Foster "On Certain Euphonic Embellishments in the Verse of Propertius" in TAPA., Vol. 40 (1909), pp. 31–62. I. HEXAMETERS (1) Monosyllabic endings: Catullus and Propertius employ them frequently; Tibullus and Ovid, very rarely. (a) Catullus has 13 examples, including pronouns, forms of esse, and forms of res. Four times his verse ends in two monosyllables. (b) Of the 31 cases in Propertius, 20 are a singular form of the first or second personal pronoun, 5 are forms of qui; 4, forms of esse; fles occurs once, and iam once. (c) Ovid in the Amores (which are used for these tests) has 4 cases, viz. a form of esse, and me, twice each. No instance (d) Tibullus (Bks. 1 and 2, which are the only safe ground for an investigation of his usage) has sint once. occurs in the book of Lygdamus. (2) Polysyllabic endings. These are more rare. They are occasional in Catullus; twice Ovid uses a quadrisyllabic proper name; Propertius has similar instances; Tibullus has none. (3) Spondees still play an important part in the hexameters of Catullus, whose taste is like that of Ennius. This appears most strikingly at the end of the verse. He has 13 spondaic verses out of 322; of these one ends in a monosyllable, one in a trisyllable, the other 11 in words of not less than four syllables. 68, 87 has 5 spondees; 116, 3 is worthy of Ennius himself, being composed entirely of spondees. In the other elegists, however, the proportion of dactyls and spondees is not unlike that of the other Augustan writers. Tibullus employs the dactylic beginning of the hexameter in the proportion of about four of these to one beginning with a spondee.1 (4) Rime. A species of middle, or Leonine, rime begins to be noted in Catullus, and continues throughout the whole group of writers, being apparently an extension, or an echo, of the very common similar rime in the pentameter. In the hexameter this rime occurs between the last syllable of the verse and that pre1Cf. Hennig, Untersuchungen zu Tibull (1905), p. 19. ceding the verse caesura, i.e. between the endings of the two parts of the verse. Not less than 41 examples of this may be found even in Catullus, e.g. 96, 1: Si quicquam mutis || gratum acceptumve sepulcris. The percentage of such cases increases in Tibullus, reaches a maximum in Propertius, and decreases again in Ovid. When this is combined with the common pentameter middle rime, and is at the same time an end rime, we have a still greater refinement, as in Tibullus, 1, 9, 25-26:— ipse deus tacito permisit lingua ministro libera verba mero.1 In many cases, though the rime is imperfect, the similarity of sounds, as of a long vowel to a diphthong, or of one vowel followed by s to another vowel and s, produces a pleasing effect, which was frequently sought by these poets, e.g. Tibullus, 2, 5, 69-70: quasque Aniena sacras pertuleritque sinu. The variety of these effects is countless. (5) Verse caesura. This depends, of course, upon the individual taste of the different authors. (a) Catullus is fairly orthodox, with 267 out of 318 hexameters exhibiting the penthemimeral caesura, 30 the hephthemimeral, 16 the feminine caesura in the third foot, and 5 the so-called "bucolic " diaeresis. One or two verses have no verse caesura at all. (b) But Tibullus, with nearly double the number of verses, shows his fondness for the hephthemimeral caesura by using it five times as often, 152 times in all, 32 times without the customary accompanying trithemimeral. A frequent added refine 1 Cf. Ovid, Am. 3, 2, 17-18; Prop. 1, 6, 17-18. |