A , )| XI* n\ 7 RM Extracted from Transactions of the American Philological Association, Vol. LI, 1920. XI. The Juvenile Works of Ovid and the Spondaic Period of His Metrical Art 1 BY PROFESSOR ROBERT S. JRADFORD UNIVERSITY OF TENNESSEE CO GD I. Introduction. Ovid and the Messalla Collection to IN his chapter on famous orators, Quintilian aptly observes That, for later generations, Cicero had become not so much name of a person as the designation of the art of oratory, may apply a similar remark to the Latin poet Ovid. To students of Roman literature Ovid means the perfected ^elegiac art, the supreme mastery of the technical side of Latin verse, to which he contributed an unparalleled ele- gance and grace. Yet it is certain that he was also a person, and we know more of the actual details of his life than of the I*- ? life of any other Roman, with the single exception of Cicero. Every schoolboy is familiar with the charming picture which ^ he has given in the Tristia of his early devotion to the Muses, ^ in spite of the remonstrances of his father and the arguments which the latter urged against unprofitable literary pursuits. X. ' Bibliography : Hultgren, Observationes metr. in poetas elegiacos Gr. et Lat., ^ Leipzig, 1871; Kone, Sprache der ram. Epiker, Minister, 1840; Radford, " Licensed Feet in Latin Verse," Studies in Honor of Maurice Bloom-field (New . V Haven, 1920), 251-272; O. F. Gruppe, Rom. Elegie, Leipzig, 1838; Klee- mann, De libri tertii carminibus quae Tibulli nom. circumferuntur, Strassburg, ^- 1876; Teuffel, Studien u. Charaktcristiken , Leipzig, 1871; Fuss, De elegiarum ^ libra quern Lygdami esse putant quidam, Miinster, 1867; Krafft, De artibus ^/ quas Tib. et Lygd. in versibus concinnandis adhibuerunt, Halle, 1874; Paroli, De Tib. arte metr. cum Lygd. comparata, Brescia, 1899; Hartung, De pane- u _ gyrico ad Messallam pseudo-Tib., Halle, 1880; Ehrengruber, De panegyr. O Messallae pseudo-Tib., partes i-x, Kremsmiinster, 1889-1899; Knappe, De 3 Tib. libri quarti elegiis, Dud -stadt, 1880; Nemethy, Albii Tibulli carmina, 3 etc., Budapest, 1905 ; Lygdami carmina, Budapest, 1906; K. F. Smith, Elegies of Tibullus, New York, 19,13; Burman, Ovidii opera omnia, Amsterdam, 1756, with Index Verborum; Eschenburg, Wie hat Ovid einzelne Warier u. Wort- klassen im Verse iierwandt?, Lubeck, 1886; H. de la Ville de Mirmont, " Le poete Lygdamus," Musee Beige, vm (1904), 339-403; Plessis, La poesie latine (Paris, 1909), 361-376 ; Schanz, Gesch. d. rom. Lit. n 3 , i, 232 ff. ; Ullrich, Studia Tib.: De libri n editione, Berlin, 1880. ^;w Vol. li] The Juvenile Works of Ovid 147 He was one of the most precocious of Roman poets, and like Cowley or like Pope he " lisped in numbers, for the numbers came." 2 According to his own account, when he first re- cited his juvenile poems to a public audience, his " beard had been shaved only once or twice." 3 This plain statement is usually misinterpreted, but clearly the natural meaning is that when he first entered the circle of his patron, Messalla, he was only fifteen or sixteen years of age. 1 Beginning to write at so early a period, when the simpler and more natural school of Catullus. Gallus, and Fropertius was still in the ascendant, Ovid passed through a long period of apprentice- ship and, after much wavering and much experimentation, eventually abandoned the more natural manner with which he had begun, and went over wholly to the more artistic and more epigrammatic style of Tibullus, which he found better suited to his own rhetorical training and to which he finally gave an undisputed supremacy in the domain of Roman elegy. It is true that a different opinion is usually held today, and it is everywhere assumed that he devoted himself from the first to the imitation of Tibullus and possessed from the beginning the remarkable facility and skill which make him easily the first of Roman metrical artists. This view- seems to me wholly erroneous, and I shall begin in the present study to trace the various stages by which Ovid, the -historical person, the friend of Messalla, the disciple first of Catullus and later of Tibullus, reached the acme of artistic perfection. As is well known, the poet Ovid possessed brilliant powers of description, and was both a natural story-teller and a - Sponte sua carmen numeros veniebat ad aptos, Et quod temptabam scribere (in prose), versus erat. Trlst. iv, 10, 2;. Ub. 5 7- 4 There was no definite time fixed for the dcpositio barbac, but it usually- coincided with the assumption of the toga 'drills (Suet. Calig. 10). Most writers incorrectly make this statement refer to Ovid's twentieth or twenty- second year, as Schanz, Gcsch. d. rom. Lit. n 3 , i, 293; J. Heuwes, De teni- porc quo Of. A mores conscripti sint (Miinster, 1883), 14; Cruttwell, Rom. Lit. 3c<>, etc. 148 Robert S. Radford [1920 highly trained rhetorician. He is noted for his " swift light- ness of touch," his brief, snappy sentences, his sharp con- trasts, and his terse and highly polished epigrams (cf. K. F. Smith, op. ciL Introd. 103 ff.). Since we seek, how- ever, to trace an actual historical development, it may be worth while to return for a moment to ' Ovid ' in the mean- ing of the perfected art of elegy, and to picture clearly the numerous refinements which his finished technique embraces. These rules include the breaking up of the long sentence and the restriction of the thought to a single distich, the avoid- ance of elision, especially in the latter half of both the hex- ameter and the pentameter, the marked preference given in the hexameter to the favorite masculine caesura (the semi- quinaria), the ending of the pentameter with a dissyllable, which should not be an adjective, but either a noun or a verb, and above all the preponderance of dactyls in the distich in direct opposition to the original and native char- acter of the language. Through the observance of these complex requirements the elegiac couplet gains an incom- parable elegance and precision, at the same moment that it loses its larger variety, freedom, and ease. Tibullus himself had narrowed the sphere of Roman elegy by rejecting the natural period and restricting the thought within the mini- mum space of two lines. This limitation, which is unknown to Catullus and Propertius, is undoubtedly a serious mistake, and it finds little justification of any kind in the previous history of the elegy. Yet, although he gave an artificial form to the distich, Tibullus, through the expression of simple and natural emotion, invariably retained its proper content, and he could not foresee perhaps that the more brilliant but more wayward Ovid would too often, through the absence of sincere and genuine feeling, merge the elegy in the epigram and make both form and content unduly artificial. Both Tibullus and Ovid then were consummate artists, but both yielded too much to the prevailing tendency of the Augustan age in restricting the liberty and the spontaneity of their Vol. li] The Juvenile Works of Ovid 149 verse. Tibullus is excessively rhetorical in form, Ovid is excessively rhetorical in both form and thought. (See also some just observations in Sellar, Horace and the Elegiac Poets, 307.) Even more characteristic of our poet, however, than the unity of the distich and the tyrannis of the dissyllabic close is the dactylic preponderance. We may say that Ovid is the poet, par excellence, of the dactylic virtuosity, which appears alike in his epic and his elegiac verse. While the Greek lan- guage is twice as rich in dactyls as in spondees, in Latin this relation is reversed, and the Roman language has almost twice as many spondees as dactyls. In all the other poets of Rome (with the exception only of Valerius Flaccus and a few genuine elegies of Tibullus' second book) the spondees considerably exceed the dactyls; Ovid alone has known - like the Medea or the Circe of his own exuberant fancy - how to transform, by the magic of his art, the slow but stately spondees of his native speech into the light and graceful dactyls of Hellenic verse. Through this supreme refinement he has brought to fulfillment the mission which Ennius, the hardy pioneer, had vaguely dreamed of nearly two centuries before, and has banished from Latium the last trace of Italian rusticity (cf. Hor. Epist. n, i, 160). He is the greatest artist in verse that Rome produced, the supreme master both in the elegy and in the epos. It is no wonder then that Lucian Miiller exclaims in ecstatic admiration (De re metrica-, 522) : " Hunc igitur virum, qui principatum haud dubie tenet artis Latinae, veneremur, hunc imitemur. hie sciat se plurimum profecisse, cui plurimum probetur Ovidius. huius quot sunt versus, totidem sunt artificia, quovis Phidiae ilia vel Praxitelis opere nonminora." 6 Our chief metrical scholars have carried their worship and adoration so far that they have ended by creating a veritable ' Ovid myth.' Ovid is Rome's one white 150 Robert S. Radford [1920 crow and he was always white. His art is free from the trammels of place and time, and knows no process of growth, no stage of development, but the Fates conferred upon the divine man 7 at his birth a faculty that was miraculously complete. Hence Lucian Miiller, op. oil. 346, pronounces a passage in the Amores, which contains several elisions, inter- polated, and declares those Epistles of the Heroines, which show one or two polysyllabic closes, to be spurious (ib. 29 ff., 259). Hultgren, op. cit. 26 ff., insists that the comparative lack of virtuosity in the de Medicamine fragment is not the result of immature art, but of pure accident, and asserts that to attribute to Ovid any work actually lacking the virtuosity would be to " place the head of Thersites upon the body of Agamemnon " (ib. 32). Ehrengruber, who has computed or collected the percentages of so many Latin poets, announces the rule that each author has his definite dactylic proportion from which he never greatly departs (op. cit. x, 12 ff.). Even such a master of statistics as Dro- bisch, when he acutely observes that " the Amores of Ovid, especially in the hexameter, fall short by a little more than 2% of the virtuosity attained in the Ars " (Ber. sacks. Ge- sellsch. xxin [1871], 33), does not seem to have grasped the full truth ; for he apparently thought of the decline in the virtuosity as evenly distributed over the entire Amores. In point of fact, nothing could be more erroneous than these views. Thus Quintilian, ix, 3, 70, had before him a collection of Ovidian epigrams quite similar to the Priapea and con- taining rather frequent polysyllabic closes ; he actually quotes the line, Cur ego non dicam, Furia, te ftiriam ? In order to determine the whole question definitely, I have carefully examined the three books of the Amores which we now possess and which were not published before 2 B.C., when the poet was forty-one. It is well known that there was an earlier edition in five books, published in 14 B.C., when the poet was already twenty-nine. According to my 7 Cf. Cic. Arch. 7, 16 : hunc dii'iiuim hominem, Africanum. Vol. li] The Juvenile Works of Ovid 151 results, nearly one-fourth of the poems in our present Amores have been retained from the first edition with little change, and still show the original spondaic form which they pos- sessed at their first publication. The mature works of Ovid, such as the Ars, show 57% of dactyls in the distich, while the first edition of the Amores, according to my reconstruc- tion, did not greatly exceed 48.5%. The remaining three- fourths of our present edition consists of poems which have either been fully revised or newly written, and if the spondaic fourth be subtracted, their percentage rises to more than $6 r , . In other words, the remaining poems show practically the same virtuosity as the mature works. In the second section of this paper I shall exhibit these results more fully and in tabular form. It is evident therefore that Ovid first com- posed in spondees, which is precisely what we should expect on a priori grounds. For, Latin being a highly spondaic language, it seems just about as possible for a youthful poet to lisp in Chinese or in Choctaw as in Latin dactyls. We may go further, however, and assert that even if Ovid, on his first appearance in 27 B.C., had possessed all the miraculous gifts with which he has been credited, he would not have actually composed in dactyls. It is true that even at this early date Tibullus was highly esteemed, 8 yet the memory of the free republic was still cherished, and the more natural school of Catullus was still preferred. Hence, in all proba- bility, even if the youthful Ovid had possessed the dactylic faculty, he would have lacked the will to virtuosity, since the latter implies also a willingness to sacrifice the larger perspicuity of expression and the normal descriptive word orders. Furthermore, two of the three highly dactylic elegies of Tibullus' second book were not composed until 22 B.C., five years later, as is shown by their imitation of Propertius (Xemethy. op. cii. .^S). The importance of the reconstruction of the first edition * . . . Leunturque Tibullus Kt placet, et iam Ir (xc. A Honshu prhiripe notus erat. Trist. n, 464. 152 Robert S. Radford [1920 of the Amores and of the consequent recovery of the youthful or ' spondaic ' Ovid cannot easily be overestimated. Freed from fanciful and unwarranted presuppositions, we are at liberty to restore the actual, historical Ovid, and we shall be able to show in the sequel, as I believe, that this great artistic genius, beginning, just like Catullus, with simple nature and therefore in some cases with only 37% of dactyls in the distich, has made in less than twenty years an unparalleled develop- ment in his art, and, by veritably creating a new language, such as Ennius and his eager successors achieved only in part, has been able, in the works of his full maturity, com- posed after the age of thirty-five, to rise to 57% of dactyls in the distich. I do not seek, however, to reconstruct the spondaic period of Ovid's art for a purely theoretical purpose nor with the aim of contrasting in an abstract manner the first Amores with all the remaining works; I wish rather to restore the poet's early life, and partly by following in the footsteps of eminent critics, such as Gruppe, Plessis, Klee- mann, Ehrengruber, Ribbeck, Marx, Schanz, and Cartault, partly by completing and enlarging their work. I hope to endow him with an appropriate set of spondaic works. These juvenile works are, from the Appendix Vergiliana. at least Catalepton, ix, and from the Corpus Tibullianum the whole of the Messalla Collection as well as two of the six elegies contained in Tibullus' second or posthumous book. The proper discussion of these poems naturally requires a series of articles. In the present study I shall deal only with the metrical development, while in several articles to be published shortly in the American Journal of Philology, I shall examine in detail, with the help of Burman's much neglected Index, the language of the juvenile poems in relation to Ovid's mature works and also sketch more fully the history of the contro- versy which has raged for more than a century among critics over the authorship and value of the Messalla Collection. 9 9 So far as concerns the Lygdamus elegies, a most just and admirable ac- count of the controversy is given by H. de la \~ille de Mirmont, op. cit. 339- Vol. li] The Juvenile Works of Ovid 153 A briefer outline, however, of the various opinions of scholars cannot be wholly omitted here, especially as the controversy over the Lygdamus poems constitutes one of the most amazing chapters in the whole history of literary criti- cism. I may begin with the so-called Messalla Collection. This latter is a collection of famous poems emanating from the household of the great Augustan statesman, Valerius Messalla. In our fourteenth-century codices it forms a single book, added to the unfinished second book of Tibullus and apparently published, like the posthumous second book, in 19 B.C. The Italian scholars of the fifteenth century have wrongly divided this additional book into two, the so-called third and fourth books of Tibullus, and for the purpose of avoiding confusion it seems best to retain this division. (Probably, as all the mediaeval citations show, there were originally only ' two books of Tibullus,' and the second book of 1 1 ii verses was first divided at the close of the Middle Ages into the second and third books which our late codices exhibit ; see Ullrich, op. clt. 69.) In the six elegies of the third book a youth apparently of eighteen or nineteen, yet already famous as a poet and employing the pseudonym of Lygdamus, seeks to w r in back the love of the fair Neaera, who has divorced him. It is probable, too, that he has given cause for the divorce, though naturally this is not admitted and he pleads a difficult case with wonderful facility. Al- though the poems of unhappy love easily admit. of being turned to ridicule by an unfair critic like Voss, I know of no juster criticism upon them than that of Teuffel, written in mature life and before he (like Dissen) had lent an ear to the slanders of Voss : " These elegies, through the freshness and sincerity of the feeling and the graceful ease of the verse, do their author no discredit." 10 As is well known, Lygdamus 403. Although the author does not fully understand the real issues, I \vish to acknowledge my deep indebtedness to the historical part of this fine study. 10 Quoted by Kleemann, op. cil. 17, from Pauly's Rcalcncyclopaedie, vi. 2, 1950 (1852). 154 Robert S. Radford [1920 tells us the exact date of his birth (5, 18). " My parents first beheld my birthday," he says, " in the year in which both the consuls met an equal fate " (cum cecidit fato consul uterque pari) he means in 43 B.C., when Hirtius and Pansa were both slain in battle before Mutina. Now precisely the same pentameter (cum cecidit, etc.) is used by Ovid, Trist. iv, 10, 6, to give his own birth-date of 43 B.C. The Renais- sance editors, Scaliger and Vulpius, perhaps attached less importance to the dozen or more whole lines and half-lines which Ovid and Lygdamus have in common, and the numerous other amazing coincidences, but they knew well that, in the case of different poets, the date of birth and the birth-line cannot be borrowed. Wishing therefore to preserve the third book for Tibullus and to prevent it from being assigned to Ovid, they pronounced the birth-line an interpolation from the Tristia and bracketed the whole distich. The same pro- cedure was followed later by Broukhuysen (1708), Wunderlich (1817), Bahrens (1876), and Ramsay (1887). Matters remained in this state until the German peasant poet of Mecklenburg and the great classical translator, Johann Heinrich Voss, appeared upon the scene (1810). Voss's works abound in homely and realistic scenes of village and country- life, and his idyl, Luise, published in 1795, furnished Goethe with a model for Hermann und Dorothea. He was naturally much attracted by the simple pastoral elegy of Tibullus, and when he came, in the third book, to the poems of Lygdamus with their brilliant pictures of elegance and wealth, he saw at once that this courtly city poet could not possibly be Tibullus. He therefore removed the name of Tibullus from the third book and substituted that of Lygdamus. So far he showed himself an excellent critic, but he did not stop here. He poured forth against the romanticist masquerading under the name of the unworldly Tibullus a torrent of vitupera- tion and of coarse abuse that is well-nigh incredible, and that many of our editors of Tibullus and many of our orthodox historians of Roman literature have ever since repeated in the Vol. li] The Juvenile Works of Ovid 155 most credulous or most servile fashion, but naturally in lan- guage somewhat more decorous and more restrained. 11 Al- though Voss never possessed any great vogue in Germany at large as a critic, his authority among classical philologians was immense, and it is chiefly to his violent attack upon the Lygdamus elegies that we must ascribe the hopeless con- fusion that has since arisen among scholars and that has made the question of the authorship of these poems an in- soluble ' mystery.' A part of the responsibility for the en- tanglement belongs, however, to W. Hertzberg, the well- known editor of Propertius. He also was possessed of great authority in his day, although his conclusions were usually hasty and Plessis, Etudes sur Properce, 80, charges him with an " incurable recklessness " (une incurable legerete). Otto Friedrich Gruppe, a critic of the first order, had just published his standard work, Die rdmische Elegie (1838). In a brilliant and forceful chapter of this book (pp. 105-143) he confidently identifies Lygdamus with Ovid and Neaera with Ovid's second wife, the " blameless spouse " of the Tristia, iv, 10, 71, who did not, however, long remain married to the poet. Hertz- berg at once came forward to answer Gruppe and, as Teuffel believed (op. cit. 380), to make the identification with Ovid forever impossible. He apparently believed that the Her aides, which stood first in the older editions, were the earliest works of Ovid. Naturally he experienced little difficulty in showing that these poems consist almost entirely of dactyls, 12 and 11 Of course every classical scholar must entertain the kindliest feeling and the greatest respect for the famous translator of the Odyssey and of the Georgics, but no account of the Lygdamus controversy can be intelligible which fails to bring out strongly Voss's violent prejudices and his scurrilous language. Fuss, op. cit. 54, cites numerous examples of the latter, and protests earnestly against such unfair methods of controversy. It appears to the present writer, however, that greater blame attaches to Voss's obsequious followers than to Voss himself. For Voss was a privileged character as a man of genius who had a distinct point of view the simple life and the return to nature. Many of the scholars, however, who have been content to take their criticism at second hand from Voss, have no special point of view and no excuse that seems valid. " In his versification Ovid skips and dances, so that his hexameter, even 156 Robert S. Radford [1920 show an excessive accumulation of rhetorical effects. There- fore, he concludes, the Lygdamus elegies, which are written in spondees and in a natural style, cannot possibly be the work of Ovid. With the recklessness which Plessis has justly noted, he adds : " These characteristics of the metre are precisely those which stand out most sharply in youth." The truth is just the reverse, and Hultgren correctly lays down the rule that the more the poet advances towards the prime of life, the more the proportion of dactyls and of dactylic beginnings increases. 13 The majority of careful students, be it said to their credit, have never accepted the prejudiced views of Voss : thus the elegies have been vigorously defended by Spohn (1819), by Golbery, the Lemaire editor (1826), by Fuss (1867), and by Cranstoun, the English translator (1872). Much more moderate views are also to be found in Cruttwell (1877), Ribbeck (1889), and Sellar (1892). Finally, Plessis, in his Pocsie Latine (1909), in the chapter devoted to Lygdamus (pp. 361-376), has at once paid a beautiful tribute to the genuine merits of these elegies and given a noble exemplifica- tion of the true critic's art. In his view, they proceed from a youthful poet of rare and brilliant genius, whose native generosity and tenderness of feeling have not yet been spoiled by contact with the corruptions of the world. Gruppe had made it probable that the Lygdamus elegies are the work of Ovid, but he had not proved it. This proof was, however, definitely rendered by S. Kleemann (1876), who, in an elaborate dissertation, with the help of Burman's in the more serious poems (Tristia and Fasti), has nothing but dactyls in almost a half of the verses " (Hallische Jahrbiicher i [1839], 1024 ff., quoted by Teuffel, op. dt. 380 ff., and by Kleemann, op. fit. 42 ff.). 13 Op. fit. 29 : " Hexametri Ovidiani illustria exempla sunt, quae decent eo magis crescere numerum dactylorum, quo magis ipse in arte procedat. . . . Dubitari amplius nequit, quin poetarum elegiacorum poemata, minus dactylice in principio distichi constructa, inter opera iuvenilis aetatis referenda, carmina autem cum plurimis initiis dactylicis florenti aetati adnumeranda sint." Klee- mann, op. fit. 29, justly lays down the same canon for Tibullus : "arte erudita in hexametris dactylus crebrior fit." Vol. li] The Juvenile Works of Ovid 157 Index, examined the language minutely and pronounced it wholly Ovidian. It should be added that Kleemann used only a part of the available material from the Index and the Ovidian corpus perhaps not more than one-half but even this, in my judgment, is far more than is strictly neces- sary for purposes of valid proof. Kleemann's study was much praised by the reviewers, but otherwise it has received little attention. 14 The reasons for this neglect are not far to seek. Many scholars had accepted without question and without independent study the distorted views of Voss which came to them in a slightly diluted form through the voluminous commentary of Dissen ; others lay entrenched in fancied security behind the barrier of the metre which Hertzberg had so conveniently and so confidently provided. Strangely enough, the Index to Ovid remained practically a closed book. As soon as the Burman Index is used, the whole Messalla Appendix is seen to be unmistakably the work of the youthful Ovid, aetate eighteen to twenty-four. 15 There is complete identity of vocabulary, and all the most characteristic Ovidian- isms are in the Appendix, except only those which were de- veloped later for the sake of the virtuosity and which are conveniently enumerated by Eschenburg. The evidence which is drawn merely from repeated tags and half-verses, without distinctive peculiarity of usage, is of course incon- clusive for an author like Ovid who borrows so freely from contemporary poets, but the proof that rests upon the plain- est and simplest idioms often recurring is one that scarcely admits of doubt or question. The Lygdamus poems, com- 11 An honorable exception is Professor K. P. Harrington's scholarly edition of the Roman elegiac poets (Selections, 1914). Professor Harrington com- ments on the possible identity of Ovid and Lygdamus with complete candor and with an open mind, although he does not commit himself definitely (Introd. 36). 15 In justice to myself, it is only fair to state that I reached my conclusions as to the identity of Lygdamus and Ovid at a time when I was acquainted neither with Gruppe's results nor with those of Kleemann and Ehrengruber, but was obliged to rely upon the simple text of Ovid and Eichert's lexicon to the Metamorphoses (Hanover, 1886). 1-8 Robert S. Radford [1920 posed at eighteen or nineteen, contain about ninety specific and striking Ovidianisms ; Kleemann gives about half of these. The Panegyric Upon Messalla contains probably more than a hundred Ovidianisms ; Ehrengruber has already called attention to all of these, and in order to explain their occur- rence, has propounded the ingenious theory that the Pane- gyric was a school exercise composed in a later age by some pupil of the rhetoricians who had access to all the works of Ovid and pilfered most freely from them all. 16 Next, as the metre shows, come the six little letters of Sulpicia, the kins- woman of Messalla, in which this lady (with Ovid's sympa- thetic assistance) undertakes the part of wooing the shy youth, Cerinthus; they contain more than twelve Ovidian- isms. About 21 B.C., aetate twenty-two, Ovid composed the five charming elegies giving in fuller form the story of the same pair of happy lovers, Sulpicia and Cerinthus ; they show more than forty Ovidianisms and 47.4% of dactyls, thus approaching closely to the proportion of the first Amores (about 48.5%). The exquisite imitation of Tibullus, Nulla tuum nobis (iv, 13), that closes the collection, has ten Ovidian- isms. Since Ovid edits the unfinished second book, the two spondaic elegies, with Ovidian language and thought, are in all probability his work, and can no longer be ascribed to Tibullus either in whole or in part. The poem in honor of Messalinus (n, 5) has, in fact, long occasioned difficulty, and has been known for nearly a century as the ' suspected elegy.' The other Ovidian elegy, n, 2, is evidently a continuation of the Sulpicia group, and celebrates the birthday of the shy but sorely smitten lover Cerinthus, who is now happily married to Sulpicia and is therefore given his true name of Cornutus. This poem is rightly assigned to the ' fourth book ' by Gruppe. The conclusions which we have reached upon grounds of language and metre are supported also by strong external 16 Op. cil. x, 71 ; n, 28; in, So, etc. This treatise of more than 700 pages in all is a magnificent collection of material and, in spite of its technically erroneous conclusion, is truly a masterly piece of work. Vol. li] The Juvenile Works of Ovid 159 evidence. Thus in the letters of the exile Ovid refers in unmistakable terms to the issuance of the Appendix, when he writes that it was Messalla who first induced him to ven- ture upon the publication of his works. 17 He mentions expressly the epicedion which he had composed upon the death of his patron and which was sung in the forum (Pont. i, 7, 29 fL), but he also refers again and again to poems which he had composed in Messalla's honor in his early youth. Thus he writes that not even the eldest son, Messalinus, can remember the time when he " first began to venerate Mes- salla " (Pont, n, 3, 79 ff.). This last statement scarcely applies, I think, to our Panegyric, but can be more fitly referred to an earlier eulogy which we fortunately still possess -I mean Catalepton ix (xi), contained in the Appendix Vergiliana, a poem which celebrates Messalla's triumph over Aquitania, and which was therefore written in the year 27 B.C. Ribbeck, Appendix Vergil. Proleg. 12, and Gesch. d. rom. Die/it, u, 200, has already, on stylistic and metrical grounds, identified the author of this elegy with ' Lygdamus,' and both Marx (Pauly-Wissowa, i, 1326, s. v. Albius) and Schanz, op. cit. 282, p. 233, speak approvingly of this view. I find on examination that these judgments are strongly supported by the language of the poem, and this elegy may therefore be confidently regarded as the earliest extant work of Ovid, written in his seventeenth year ; in fact, in a brief monograph and admirable commentary which was published some years ago, but has just come into my hands at the moment of writing (De Ovidio elegiac in Messallam auctore, Budapest, 1909), Nemethy, I find, has already clearly per- ceived and, in large measure, convincingly demonstrated the Ovidian authorship of the Catalepton. Messalinus, the eldest son, quindecemvir 19 B.C., consul 3 B.C.. was probably born about 38-36 B.C., and can therefore scarcely have been more than ten years of age when this eulogy was composed. I have not yet spoken of the date of the second eulogy 17 Pont, u, 3, 75 ff. (addressed to Cotta, the younger son). 160 Robert S. Radford [1920 which we possess, namely, the elaborate Panegyric of the Messalla Collection. This brilliant and highly rhetorical work is metrically more advanced than the Lygdamus elegies and was certainly composed at a later date than these poems. Undoubtedly, by an ingenious literary or artistic fiction, the Panegyric itself purports to be written in the year of Messalla's consulship, 31 B.C., and the events of the years 30-27 B.C., namely, the expeditions to Aquitania and the Orient, which had already been expressly celebrated by Tibullus (i, 7) and by the youthful Ovid (Catalepton, ix) , are nowhere mentioned as having actually occurred. Special students of the Pane- gyric, however, have long seen that the Gallic and Egyptian campaigns are well known to the clever writer, and are most skilfully introduced into the poem by way of prophecy (vati- cinium ex eventu}. 18 The Panegyric was therefore composed after the Lygdamus elegies, and the first draft of the poem must have been drawn up about the year 23 B.C. ; it treats chiefly the earlier career of Messalla, which had not previously been made the subject of poetic encomium. We are not at present fully in a position to state how Ovid was occupied in the interval between the composition of the Lygdamus poems and the Panegyric. It is true that H. de la Ville de Mirmont confidently assumes (Jeunesse d'Ovide, 209) that it was shortly after the two early marriages and about his twentieth year that our poet visited Asia Minor and Sicily in the suite of the poet Macer, and at first glance the Pane- gyric also, in its present revised and perfected form, appears to contain probable or possible references to Sicily (vss. 197, 200). We must carefully refrain, however, from drawing hasty conclusions at present and must frankly admit that we cannot at once determine the exact date of the year which Ovid spent with Macer in Asia and Sicily (Pont, n, 10, 21 ft".). A brief word must be said, however, upon Ovid's relation to the Ciris (the legend of Scylla and Nisus) in the Appendix 18 See Hartung, op. ell. 38 ff. ; Ehrengruber, op. cit. i, 7 ; x, 71 ; also Belling, Albius Tibullus: Untersuchung, 205 ff. Vol. li] The Juvenile Works of Ovid 161 Vergiliana. As is well known, this poem offers special peculi- arities and difficulties. Besides being partly a translation from Greek sources, it everywhere closely follows the manner of Catullus and, owing to the poet's prodigious memory, in many passages it presents almost the appearance of a cento compiled from Catullus and Vergil. It was composed after the publication of the complete Aeneid (19 B.C.), the whole of which it imitates. The agreement therefore of the Ciris with the usual Ovidian vocabulary is not quite so close as we find in the other juvenile works, yet it is sufficient, I believe, easily and conclusively to establish Ovidian author- ship, especially when we consider that, by a species of /eeW> =T X'^N^ V "^ y i = ^s '' ^MiHRAHY,-9/: | 11(7 i Sc^i o X....^' vV */, } I "Ti i'L^Cil < l -TN > Ti ! J * ' V:-,: UC SOUTHERN REGIONAL LIBRARY FACILITY A 001 410597 7