STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. THE SOURCES AND THE EXTENT OF PETRARCH'S KNOWLEDGE OF THE LIFE OF VERGIL By Duane Reed Stuart A conspicuous feature of Vergilian study in the last decade and a half has been the effort to arrive at a saner conception of the events of the poet's life than has held good hitherto. That substantial progress has been made in this direction is undeniable. The higher criticism of the Appendix and of the Eclogues which, at least from the beginning of this century, has usurped perhaps the supreme place in the interests of students of Vergil has precipitated fruitful, though not always final, discussion of various topics germane to the poet's life and literary activities and has furnished us with new standards for differentiating Wahrheit und Dichtung in his works. Further- more, thanks to the editions of the Vitae Vergilianae produced by Diehl 1 and Brummer, 2 books which supplied a long-felt want, and to sundry articles from the pens of Klotz, 3 Kroll, 4 Leo, 5 Norden, 6 Vollmer, 7 and others, we may boast of a newly acquired intelligence in the use of the ancient biographies of Vergil. These additions that latter-day research has made to our knowl- edge of Vergil's life have been due, of course, not to the discovery of new documentary evidence, but to the rationalistic criticism of sources long existing. An enlightened estimate of the reliability of the data contained in these sources has freed us from inveterate and "vulgar errors." So far as the raw material for the reconstruction of Vergil's life is concerned, the humanists of the fifteenth century were equipped about as adequately as are we. At least, after the discovery of the Catalepton — as Birt has taught us in Jugendverse 1 Die Vitae Vergilianae und ihre antiken Quellen (Bonn, 1911). 2 Vitae Vergilianae (Leipzig, 1912). »Rh. Mus., LXVI (1911), 155-60; LXVII (1912), 306-9. * Ibid., LXIV (1909), 50-55. • Hermes, XXXVIII (1903), 1-18. •Rh. Mus., LXI (1906), 166-177. ' Ibid., pp. 481-90; SB. der Bayer. Ak. Phil.-Hist. Kl., 1909, Abhand. 9, pp. 6-11. [Classical Philology XII, October, 1917] 365 366 Dttane Reed Stuart und Heimatpoesie Vergils, no negligible source of information — they could have laid their hands on all the essential primary sources included in our modern apparatus. Nolhac goes so far as to say (Petrarque et I'humanisme, 2d ed., I, 124) that even Petrarch, at the dawn of the Revival of Learning, knew concerning the life of Vergil all that can be learned from the poems and well-nigh all that the ancient sources and testimonia have to offer. Petrarch's acquaintance with the genuine version of the Vita written by Aelius Donatus and based, as we know, on the biography of Vergil included in the De viris illustribus of Suetonius has long been regarded as indisputable. Nevertheless, no formal attempt has been made to present the evidence on which this belief rests except by Sabbadini, 1 whose studies have contributed in large measure to our knowledge of the classical scholarship of the Renaissance. Nolhac in both editions of his great work asserts categorically that Petrarch had read the Life of Vergil written by Donatus, but only in the first edition does he adduce any evidence in support of his affirmation. This evidence consists in a single datum which, unfortunately, turned out to be quite untenable. In note 6, p. 106, Nolhac reported the existence on the cover-leaf of the celebrated Vergil of Petrarch, which is now one of the treasures of the Ambrosian Library, of a long citation accompanied by the surface reference, Donatus in vita Virgilii. As Sabbadini subsequently pointed out, 2 the citation is not derived from the Suetonian Life, and the supposed surface reference is non-existent. Petrarch's appeal is to the com- mentary of Donatus on the Eclogues, whatever it was that the human- ist knew under this name. 3 Nolhac tacitly admitted his mistake by excising this part of his original note from his second edition (see p. 124). However, he left the statement in the text unaltered, presumably because he was content to rest his case on Sabbadini's 1 "Quali biografie vergiliane fossero note al Petrarca," Rend, del r. ist. Lomb. di sc. e Hit., XXXIX (1906), 193-98. * Op. cit., p. 197. ' Sabbadini, Le scoperte dei codici lalini e greci, pp. 38-39, presents evidence that Petrarch had a commentary on the Eclogues, believed by him to be the work of Donatus. This is another rubric in the enigmas attached to the history of the commentary of Donatus. Let us hope that Professor Rand has put us on the road to a solution; see Class. Quart., X (1916), 158. Petrarch's Knowledge of the Life of Vergil 367 article, which had appeared just prior to the publication of the second edition of Pitrarque et I'humanisme. As the result of an endeavor to test for my own satisfaction and instruction the validity of the arguments advanced by Sabbadini in favor of the accepted view, I have found myself, to my own surprise, forced to the conclusion that the treatment previously accorded to this topic is unconvincing and incomplete. The pertinent data are far from being exhausted. I have tried, therefore, to assemble with greater fulness than has been done all that Petrarch knew or thought that he knew about the life and the personality of Vergil. Thus we shall be in a position to project upon the point at issue all available evidence and to realize in general the character and the sources of the great humanist's information concerning the Roman poet who was one of the chief idols and mentors of his intellectual life. First, I shall have to crave the patience of the reader while I summarize Sabbadini's arguments in the order in which he sets them forth. 1. In Petrarch's text of the Servian commentary on the works of Vergil, which frames the text of the poems in the Ambrosianus, Tarentum is cited as the place where Vergil breathed his last. 1 According to Donatus, 11. 131-32, 2 and Jerome Euseb. Chron. 2. 1998, the poet died at Brundisium. Petrarch twice refers to this variant tradition, once in Famil. 13. 4, Vol. II, p. 223, Fracassetti: "cuius cinerem vel Tarento ereptum vel Brundisio tua possidet ac sua Parthenope," again in Itiner. Syriac: "Tarentum tibi monstrabitur .... fatalis locus, quamvis alii Brundisium dicant." In the latter reference, alii, says Sabbadini, would seem to point to Donatus and Jerome. 2. In the Life by Donatus, 11. 100-103, is found the allusion, quoted by all the handbooks, to the pre-eminence bespoken for the Aeneid by Propertius in the famous couplet 2. 34. 65-66: "Aeneidos vixdum coeptae tanta extitit fama, ut Sextus Propertius non dubi- taverit sic praedicare: cedite Romani scriptores, cedite Grai: nescio quid maius nascitur Iliade." > The passage in question is found in the later manuscripts of Servius, such as the Codex Dresdensis; see Thilo-Hagen, I, p. 4. » Unless otherwise indicated, references to the Vita Donatiana are given according to Brummer's text. 368 Duane Reed Stuabt In Petrarch's second letter to Cicero (Famil. 24. 4, Vol. Ill, p. 266 Frac), we read: "ut Aeneidi cederet Ilias [te] iussurum fuisse non dubito, quod iam ab initio Vergiliani laboris Propertius asseverare non timuit. Ubi enim Pierii operis fundamenta contemplatus est, quid de illis sentiret, et quid speraret aperte pronunciavit his versibus : cedite," etc. Sabbadini asserts that Petrarch's words could not have been based directly on the text of Propertius, but must have been suggested by the context in the Vita, since here alone do we find definite, chronological allusion to the inchoate state of the Aeneid, expressed in the sentence "vixdum coeptae .... fama," to which the phrase "iam ab initio Vergiliani laboris" harks back. 3. In Famil. 3. 11, Vol. I, p. 164 Frac, Petrarch refers to letters written by Augustus to Vergil and Horace, and to the democratic tone adopted by the Princeps in this correspondence with his two friends : "Mitto alia, iis fortasse mirabilia, .... qui .... principis illius ad eosdem humiles amicos, velut ex aequo missas, et saepe dulcibus blanditiis refertas, epistolas non legissent." Sabbadini remarks that only in the Life by Donatus is reported the letter of Augustus to Vergil. The context is given in 11. 104-7 : " Augustus vero . . . . supplicibus atque etiam minacibus per iocum litteris efnagitaret, 'ut sibi de Aeneide,' ut ipsius verba sunt, 'vel prima carminis vroypaxfrri vel quodlibet icuiXov mitteretur.' " 4. Near the end of the poetical epistle to Vergil, Petrarch refers to the frustration of Vergil's wish that the Aeneid be burned: .... miserum Aeneam iam summa premebant fata manu, iamque ore tuo damnatus abibat, arsurumque iterum pietas Augusta secundis eripuit flammis, quern non morientis amici deiecti movere animi, meritoque supremas contempsisse preces aevo laudabitur omni [Famil. 24. 11, Vol. Ill, p. 292 Frac] Sabbadini sees in the words arsurumque .... flammis a reminis- cence of the celebrated epigram of Sulpicius, quoted in one of its two variant versions in the Life, 11. 142 f. The concluding couplet runs thus: infelix gemino cecidit prope Pergamon igni, et paene est alio Troia cremata rogo. Petrarch's Knowledge of the Life of Vergil 369 Now, in my opinion, the cogency of these arguments is by no means absolute; examination will show that they do not all stand close scrutiny. Certain objections are bound to occur to one who is intent on settling beyond peradventure of a doubt the interesting, and not entirely trivial, question as to the biographies of Vergil known to Petrarch and utilized by him. Apropos of Sabbadini's first point, it may well be urged that the plural alii does not definitely betoken a duality of sources. It is quite conceivable that Petrarch might have derived from the Chron- icon only the tradition that makes Brundisium the scene of Vergil's death, and still have permitted alii to slip from his pen. Such laxity in reference is, as we know, one of those venial sins against the ethics of strict quotation that find ample analogy in the writers of every age. Or by the plural Petrarch may have intended to allude, not specifically to variant tradition found in one group of literary sources, but in general terms to report the currency, among one set of the litterati and curiosi of his own time, of the belief that Brundisium had stronger claims on the melancholy distinction than had Tarentum. Luckily, however, we do not have to confine ourselves to con- jectures as to what liberties Petrarch may have allowed himself in the technique of quotation. There is a third passage, overlooked by Sabbadini, the evidence of which is directly pertinent. In De remediis utriusque fortunae, Lib. 2. dial. 125 (p. 686, edition of 1649), Petrarch writes: "Virgilium mundo dedit Mantua; Brundisium, sive ut alii perhibent, Tarentum rapuit, nunc Neapolis tenet." Now, as I have previously remarked, the statement that Vergil died at Tarentum is found in certain late manuscripts of Servius, in which we find appended to the Servian biography of the poet the words: "periit autem Tarenti in Apuliae civitate," etc. (see Thilo- Hagen, p. 4; Diehl, p. 42). The exemplar from which Petrarch's text of Servius was derived belonged to this class. Furthermore, the Servian Life is the only literary source that presents this tradition. Hence it is evident that in the passage cited above Petrarch used alii loosely or was referring primarily to contemporary opinion. Evi- dently no inference of any weight can be derived from his use of alii in the passage from the Itiner. Syriac. cited by Sabbadini as an 370 Duane Reed Stuart indication that Petrarch had in mind the Life by Donatus as one of two sources. As to Sabbadini's second argument: only a rather determined prejudgment could maintain that the lines of Propertius, read in the original context, give no pretext (non da alcun appiglio) for such chronological assumptions as those expressed in the words of Petrarch, "iam ab initio Virgiliani laboris" and "operis fundamenta contem- platus est." The context in Propertius speaks quite plainly of the inchoate Aeneid: 61 [iuvet] Actia Vergilium custodis litora Phoebi Caesaris et fortes dicere posse rates, qui nunc Aeneae Troiani suscitat arma iactaque Lavinis moenia litoribus. Cedite, Romani scriptores, cedite Grai: 66 nescioquid maius nascitur Iliade. Lines 63-64 contain an unmistakable reminiscence of the opening lines of the Aeneid, as Rothstein ad loc. has observed. Certainly no great exegetical acumen was essential in order to discern that it is the initia ipsa of the Aeneid that are here echoed, that "nunc Aeneae Troiani suscitat arma" "is now stirring to action the arms of Trojan Aeneas," means in poetical but plain language that the Aeneid was in the early stages of composition. Nascitur also is no obscure chronological indication, whether read in the context of the elegy or in the isolated couplet. We know enough about the technique of the author of the Vita to realize that probably the sole authority for the words Aeneidos vixdum coeptae was inference based on the passage from Propertius. It would be a needlessly harsh commentary on the intelligence of Petrarch to deny him the possession of the same degree of critical sense. Petrarch had read Propertius directly; two citations in the Ambrosian Vergil prove this fact beyond cavil. 1 It is, of course, one thing to contend that Petrarch's allusion could have been inspired solely by his knowledge of the elegy of Propertius, another to maintain that this was actually the case. One's attitude will be directly affected by what one can learn as to whether Petrarch knew the Life by Donatus and how far he depended 1 See Nolhac, I, 171; on reminiscences of Propertius in the poetry of Petrarch, consult J. S. Phillimore's addendum to Ellis, Catullus in the 14th Century (London, 1905), pp. 29-30. Petrarch's Knowledge of the Life of Vergil 371 on it, if at all. We must therefore reserve decision until all the forthcoming evidence has been examined. In the meantime, it is none the less essential to sound criticism to call attention to the over- dogmatism of Sabbadini in his treatment of this particular point. There is an additional possibility which no thoroughgoing dis- cussion should fail to take into consideration. Sabbadini assumes that only an alternative of sources existed for Petrarch's knowledge of the couplet, that he had perforce read it, if not in the elegy directly, then in the Vita. Herewith the fact is overlooked that the couplet has a textual tradition apart from both Propertius and the Vita. As the estimate given by one great poet of the masterpiece of a greater contemporary, the lines naturally attained great celebrity, hence were taken up by the Florilegia. Thus they are found in the Codex Salmasianus and in Parisinus 8069; see Baehrens, PLM, IV, 158, No. 163, Riese, Anth. Lat., I, 214, No. 264. Certain poems of the Anthology were known to Petrarch — of one of these more hereafter. We cannot, of course, presume to say that a sylloge containing this couplet had been in Petrarch's hands. It may, however, with propriety be asserted that even a limited currency of the distich in extant excerpta enhances the possibility of Petrarch's having read it elsewhere than in the Life. A denial of the truth of one of Sabbadini's alternatives therefore does not inevitably prove the other. As it appears in the Anthology, the titulus indicates the authorship of the distich, and, as has been said, nascitur alone would justify inference on the part of an enlightened exegete. Sabbadini's third argument, viewed by him as un altro indizio sicuro, does at first sight seem to be the most conclusive of all. He takes it for granted that the passage in Famil. 3. 11 transcribed above, in so far as it relates to the correspondence of Augustus with Vergil, was evoked by acquaintance with the brief quotation culled by the author of the Vita from the letter to Vergil. It is true that only in the Life by Donatus among our extant sources does an excerpt from this letter occur. Priscian, an author studied by Petrarch from his youth on and frequently cited by him, quotes (p. 901, Keil, Vol. II, p. 533), the words excucurristi a Neapoli addressed by Caesar to Vergil and, as we must suppose, notwithstanding the absence of surface reference, extracted from a letter written by the 372 Duane Reed Stuart Princeps to Vergil while the poet was sojourning in Campania. These three words offer scant grounds for such a generalization as Petrarch's, yet they do show that Augustus wrote to Vergil as man to man. From Macrobius also Petrarch knew that Vergil was a favored correspondent of Augustus; in the oft-quoted passage, Saturnalia 1. 24. 10-11, he had read a part of Vergil's reply to the request of Augustus for some specimen portion of the Aeneid, a petition which was preferred in the letter partly quoted in the Life. Vergil's reply begins: "ego vero frequentes a te litteras accipio." Petrarch, of course, was intimately acquainted with the works of Macrobius. How well he knew this passage in the Saturnalia is shown by Seniles 4. 5, p. 785, Opera, 1 where he alludes directly to the letter of Vergil, reproduces the substance, and echoes the language of the whole context, thus: "Virgilium .... constat divino illo in opere quod sibi ultimum est ... . altius aliquid sensisse quam quod loquitur, idque non modo poetantium communis habet opinio sed poetae ipsius epistula quaedam ad Augustum Caesarem scripta testatur, ubi se rem maximam et, praeter id quod apparet, studii multiplicis indigam dicit ingressum." Compare the words of Macrobius: "si in hac opinione es ut Maro tibi nihil nisi poeticum sensisse aesti- metur .... audi quid de operis sui multiplici doctrina ipse pro- nuntiet"; and infra: "ut . . . . tantum opus ingressus mihi videar." We must not, of course, forget that, in the passage cited by Sabbadini, Petrarch mentions the possibility of reading letters written by Augustus to his two poet friends. The same implication is contained in other passages in Petrarch. Petrarch is noticeably fond of reverting to the topic of the close relations maintained by Augustus with the two men of genius who were so far beneath him in the social order. Naturally the imagination of the neo-Latin bard, himself an intimate of the princes of his day, was impressed by the amiability, not to say eagerness, with which the erstwhile lord of the world cultivated the friendship of the two great poets of Rome and strove to place the relationship on a footing marked, not only by absence of all condescension on his part, but by a display 1 In conformity with the usual practice, references to the Opera omnia are cited from the second edition issued at Basle in 1581. Petrarch's Knowledge of the Life of Vergil 373 of democratic informality and self-depreciation. We may add to the passage cited by Sabbadini the following allusions to the corres- pondence: Famil. 23. 2, Vol. Ill, p. 184 Frac: "si enim Virgilio, si Flacco gloriosum fuit Augusti Caesaris et notitiam et convictum et epistulas promereri"; Senil. 11. 1, p. 881, Op.: "Augustus Caesar cui temporali potentia par non fuit, saepe humilibus cum amicis, nominatim cum Virgilio atque Horatio poetis, altis 1 viris ingenii sed ortus humillimis, 1 familiarissimas fecit epistulas ut dulce sit mirumque legentibus in tam sublimi statu talem animum tantamque man- suetudinem inveniri"; Senil. 14. 1, Fracassetti's Italian translation: " De re publica optime administranda liber " (p. 385, Op.) ; " [Augus- tus] habuit ergo in sodalicio .... Publium quoque Virgilium, Horatium Flaccum, poetas egregios, ad quos sunt ipsius principis epistulae, quibus ille summus hominum mundi dominus duobus illis rusticanis, Mantuanae ac Venusinae originis, non se aequat tantum- modo sed submittit; De remediis vtriusque fortunae"; 2. dial. 5: "vester ille Virgilius rusticanis parentibus fuit nee se libertino ac praecone patre natum Flaccus erubuit; .... Ipse [Augustus] horum duorum ignobilium .... amicitiam et convictum ceu grande aliquid blandis ac dulcibus epistolis fiagitaret." Now Petrarch's characterization of the tone and spirit of the missives of Augustus is, strictly speaking, applicable only to the content of the fragments of the Princeps' letters to Horace preserved in the Suetonian Vita of this poet. To these Petrarch's comments are peculiarly apposite. Submittit is a word perfectly describing the humility displayed by Augustus in his overtures. It is only in the letters to Horace that we find the Princeps "wheedling in honeyed phrases for friendship and camaraderie." Petrarch was undoubtedly familiar with the Vita of Horace. In Famil. 20. 7, Vol. Ill, p. 26 Frac, he quotes from it almost verbatim a sentence of the letter written by Augustus to acknowledge the receipt of one of the books of Horace and voicing a jocular protest on the part of the Princeps against the brevity of the libellus. 2 In themselves, the few words 1 Thus the edition of Basle; presumably we should read alii .... humillimi. '"Vereri autem mihi videris ne maiores libelli tui sint quam ipse es," Roth, p. 298; cf. Petrarch, op. tit.: "non tuam, fateor, legere visus epistolam sum, sed Flacci, eum quo ioeans Caesar: 'vereri,' inquit, 'mihi videris ne maiores res libelli tui sint, quam ipse es.'" 374 Duane Reed Stuart quoted by Donatus from the letter to Vergil give no warrant what- ever for inclusion in the characterization so accurately given of the spirit pervading the letters to Horace and of the phraseology in which these are couched. There is nothing but Donatus' comment on the vein of the letter to Vergil — "supplicibus atque etiam minacibus per iocum litteris" — to give pause to an expression of utter skepticism as to whether Petrarch's several allusions were founded at all on a knowledge of this section of the Life. Certain it is that if he knew this context his recollection of it became so merged in the more vivid impression naturally exerted by the racy and comparatively extensive fragments in the Life of Horace that he could forget that in the biography of Vergil there is no entreaty made for intimate friendship, but simply for advance sheets of the Aeneid. Should we not, therefore, admit the possibility that Petrarch's comments may be the result of generalization based on the passage in Macrobius and on our extant fragments of the letters to Horace, granting, perhaps, that the passage in Priscian partakes too much of the character of the proverbial needle in a haystack to make probable the assumption that it also may have played a part in crystallizing Petrarch's ideas on this subject? The objection that the implica- tion in more than one of the passages quoted above is to the effect that letters, or at least a letter, written by Augustus to Vergil, were accessible to Petrarch and had been read by him, will appear by no means final to one who recalls what we know concerning the workings of the humanist's imagination, how tricks of memory and lapse of time conspired to make him the victim of self-deception as to the works of ancient literature which he had read. Such hallucinations were furthered by the bibliophilic conditions of the times; any day might be marked by the discovery of some new treasure, the work might be handled or possessed by scholar and collector, then, through such causes as petty proprietary jealousy, unscrupulous "borrowing" — to adopt Huck Finn's euphemism for a "shorter, uglier word" — might vanish into the limbo of the inaccessible, there to remain, perhaps, for years. The most famous instance of the expansiveness of which Pet- rarch's imaginative processes were capable in respect to the events of his intellectual life is to be seen in the development into an idSefixe Petrarch's Knowledge of the Life of Vergil 375 of the notion that he had once possessed the treatise of Cicero, De Gloria. Nolhac's reasoning on this subject (I, 260 f.) is most convincing. But we do not need to confine ourselves to one hypoth- esis, however attractive, to find support for another. The workings of an analogous obsession may be traced in connection with this very point that here concerns us, viz., the basis of Petrarch's apparently circumstantial knowledge of the correspondence of Augustus with Vergil. Nolhac (I, 266, n.) has called attention to Petrarch's assertion, Res. memor. 1. 2, p. 395 Op., that, in his youth, he had once had in his hands a book of epigrams and letters composed by Augustus, but that in subsequent years he had sought the work in vain. The contents are definitely appraised: "scripsit et epigrammatum librum et epistularum ad amicos, conditum facetissima gravitate et lucu- lentissima brevitate"; also certain external characteristics are noted: "opus inexplicatum et carie semesum." To be sure, Franz Riihl in B.Ph.W., 1895, col. 468, suggests that perhaps this is not a case of confusion on Petrarch's part, but that he had actually seen such a book, identical, perhaps, with a Liber Octaviani imperatoris, listed in a twelfth-century catalogue from Limoges. 1 It is very unlikely, however, that Petrarch had ever seen, though he came to believe that he had, a book of such a nature from the pen of Augustus. It is barely possible that some late compiler had assembled from such sources as the biographies of Augustus, Tiberius, Claudius, written by Suetonius, and the Vita Horatii the fragments of Augustus' correspondence, and that this was the book that Petrarch had once possessed for a time. It is far more likely that the words in Res. memor. resulted from a confusion in Petrarch's mind of the Liber epigrammatum, mentioned by Suetonius (Vit. Aug. c. 85) as among the poetic attempts of the Princeps, and of the numerous quotations from the letters of Augustus found in the biographer. As to Ruhl's theory of a genuine Liber Octaviani existing down to humanistic times, it may be remarked in passing that such a combination by Augustus, into one book, of poetic epigrams and prose epistles, as the literal acceptance of Petrarch's words entails, would have been an incredible breach of the conventionalities of literary arrangement. 1 See Manitius, Philologisehes aus alten Bibliothekskatalogen, Rh. Mus., XLVII (1892), "Erganzungsheft," p. 27. 376 Duane Reed Stuart A piece of kindred evidence, germane to our question and quite indubitable, is forthcoming from the following sentence, Famil. 24. 12, Vol. Ill, p. 294 Frac. : " Virgilii epistolas oratione libera non inamoenas legi." As a matter of fact, Petrarch could have read only the fragment of the letter preserved in Macrobius. For the moment, this fragment is multiplied into a plurality by an exag- geration, psychic or verbal, it is difficult to decide which. The explanation which best accords with the line of indications that we have been following is to assume the influence of an aberration of the memory. In either case we have an unquestionable proof that great caution must be exercised in basing on what Petrarch tells us about the correspondence of Augustus and Vergil conclusions as to the sources from which this apparent knowledge was derived. In the presentation of his fourth argument, Sabbadini's method is not characterized by the acumen which usually attaches to his work. It is an erroneous assumption that the echo in the poetic Epistle to Vergil, if echo it be, must necessarily go back solely to the hexastich commemorating the proposed burning of the Aeneid and incorporated, with mention of its author, Sulpicius, in the Life by Donatus. The conceit that, figuratively speaking, a second Iliupersis was averted when the Aeneid was saved from the flames is a favorite topos of the metrical — I shall not say poetical — Ivdi and epigrammatic tours de force which were evoked by the name and fame of Vergil, and which are plentifully illustrated in the Latin Anthology. The motif, for example, is expressed perhaps more graphically in a variant of the hexastich. I quote from Riese, Anth. Lai., II, 121, No. 653; Baehrens, PLM, IV, 169, No. 177, took unwarrantable liberties with the text. Carmina Vergilius Phrygium prodentia Martem Secum fatali iusserat igne mori. Tucca negat, Varius prohibet, superaddite Caesar Nomen in Aenea (Aeneae Bticheler, recte ?) non sinis esse nefas. O quam paene iterum geminasti funere funus, Troia, bis interitus causa futura tui. The theories voiced in the last century by Baehrens (Praef., pp. 44—45) and by others as to the origin and the authorship of the two versions, a question complicated by the fact that in the so-called Petkarch's Knowledge of the Life of Vergil 377 Life by Probus the first four lines of the epigram, in the version appearing in Donatus, are attributed to a Servius Varus, otherwise non est, need not concern us here. The suggestion of Norden (Rh. Mus., 61 [1906], p. 175, n. 1) that Servius Varus is a stupid invention, and that the same Sulpicius played one tune on two keys, as we know the Greek epigrammatists and Martial sometimes did, is a rational explanation of the existence of the doublet. For us, as it should have been for Sabbadini, the significant thing is that there existed, apart from the text of the Life by Donatus, a version of the hexastich which, so far as its language goes, furnishes as close a parallel to Petrarch's "Aenean arsurumque iterum .... secundis flammis" as does the variant found in the Vita. To be sure, the version in the Anthology is found in one manuscript only — i.e., E — a fact which of course makes against the suggestion that it was this version with which Petrarch was acquainted. None the less, in the interests of accuracy Sabbadini should have apprised his readers of the existence of any possibility that might modify his conclusion. The theme of the second burning of Troy occurs also in another poem of the Anthology, one which possessed a somewhat wider textual dissemination, i.e. PLM, IV, 183, No. 185 = Riese I, 198, No. 242. In the first four lines the death of Vergil and his injunction that the Aeneid be destroyed are related; then the piece continues: Roma rogat, precibusque isdem tibi supplicat orbis, Ne pereant flammis tot monumenta ducum. Anne iterum Troiam, sed maior, flamma cremabit ? This poem is found in the Codex Salmasianus under the heading "Unde supra," which looks back to some other poem treating a like theme. It is also contained in the Vatican manuscript No. 1575 (Z Baehrens) of the eleventh century, with a heading which, with greater piquancy than regard to chronology, attributes it to Cornelius Gallus. This fiction was doubtless instrumental in securing to the decastich the popularity which it evidently came to possess, for it is found rather widely distributed in manuscripts of the fifteenth century as well as in the older editions of Vergil. These instances of the motif, however, are of minor significance for our purpose in comparison with another which I am about to add; they have been cited in order to show that Petrarch and his generatio 378 Duane Reed Stuart might conceivably have met the theme in sources other than the Life by Donatus, a contingency which Sabbadini totally ignored. We need not, however, confine ourselves to indicating possibilities. We can, if I am not oversanguine, lay our hands on the specific influence to which Petrarch reacted in penning the lines in the Epistle to Vergil. I refer to the famous little poem in the Anthology (Riese I, 2, No. 672; Baehrens, IV, No. 183) the authorship of which is attrib- uted to Augustus by the manuscripts. The first three lines are: Ergone supremis potuit vox improba verbis Tam dirum mandare nefas ? ergo ibit in ignes Magnaque doctiloqui morietur Musa Maronis ? In 11. 27f . we find again the conceit developed as follows : Si mens [i.e. Vergilii] caeca fuit: iterum sentire ruinas Troia suas, iterum cogetur reddere voces 1 ? Ardebit miserae narratrix f ama Creusae ? Sentiet appositos Cumana Sibylla vapores ? Uretur Tyriae post funera vulnus Elissae Et iurata mori, ne cingula reddat, Amazon ? We know that Petrarch was acquainted with this poem. Sab- badini, op. cit., p. 197, Nolhac, I, 125, n. 1 and, on the authority of these scholars, Cosenza (Petrarch's Letters to Classical Authors [Chicago, 1910], p. 146) cite two allusions. One of these is contained in a marginal note in the Ambrosianus apropos of the words of the Servian Life, "hac lege iussit emendare," Thilo-Hagen, I, 2, 13; Petrarch commented as follows: super hoc elegantissimo carmine se excusans." Again, in Opera 3, p. 90, there is written: Carmen adhuc superest quo Caesaris atque poetae Maiestas studiumque vigent semperque vigebunt. To these two references should be added Res memor. 1. 2, p. 396, Op.: " nee poeticae expers fuit [Augustus] ; extat eius in Virgilium carmen breve quidem sed nee humile nee insulsum." This third citation, previously overlooked, has, for a reason which will soon appear, an importance not shared by the other two. Now, in the humanistic redaction of the Life by Donatus (Diehl, p. 33, sec. 58; Brummer, p. 28) 11. 1-3 and 18-22 of the poem are quoted. Sabbadini observes that the manner of Petrarch's references would indicate that he knew 1 Poenas, Vat. 1575, Baehrens; lucem vel luces'! Riese. Petrarch's Knowledge of the Life of Vergil 379 the poem as a whole, hence must have read it in some one of the collections of excerpts now gathered into the Latin Anthology. I am glad to indorse this opinion, although I do not place great stress on the tone of Petrarch's citations, in so far as this is exemplified by the two contexts mentioned by Sabbadini and Nolhac. In the first instance, it is only an appreciation of the literary merit of the poem that is expressed; this could have been founded on a knowledge of a few lines merely. Furthermore, it must be confessed that it would be quite in consonance with his method of expression to imply that he knew the entire poem, even if he had read only extracts from it. We have seen that he writes as if he had read complete letters of Augustus to Vergil and Horace when, as a matter of strict accuracy, he could at the most have read excerpts only; that he so refers to the letters of Vergil that an unwary reader might surmise that he had read a collection instead of a fragment of one. There are weightier reasons which justify the conclusion that Petrarch knew the poem as a whole and hence derived his acquaintance with it from some other source than the interpolated version of the Vita. (1) The existence of the Donatus Auctus in its present form cannot be traced back beyond the first quarter of the fifteenth century; 1 Nolhac's assertion (I, 124) that Petrarch probably knew the interpolated version cannot be substantiated. As I shall hereafter point out, there is only one instance of parallelism even suggesting dependence by Petrarch on the Auctus, and this instance will not stand. (2) The popularity of the poem was great and it was widely disseminated. How numerous the manuscripts are in which it is incorporated a glance at the apparatus cited by Baehrens and Riese in their editions will show. (3) In the passage from Res memor., to which attention is called above, the word breve, unlike references to the artistic character of the poem, certainly seems to betoken knowledge of the poem in its entirety. It is indeed surprising that Sabbadini should assume on one and the same page that Petrarch knew the poem Ergone supremis, etc., in its entirety, and should also argue that the words in the Epistle to Vergil actually echo the epigram of Sulpicius. Evidently 1 Sabbadini, he biografie di Vergilio antiche medievali umanistiche, Studi Ital. di filol. class., XV (1907), 260. 380 Duane Reed Stuart 11. 27-32 of our poem, which have been quoted above, must have escaped Sabbadini's memory, else their claims to consideration as the source of Petrarch's inspiration would not have been overlooked. Credit should be given to Cosenza for avoiding this pitfall. In his note on the lines in the epistle (op. tit., pp. 145-46) he rightly asserts that Petrarch knew the story of the rescue of the Aeneid "also from the famous poem Ergone supremis," as well as from Macrobius and Donatus, and regards the poem as first in order of importance among the three sources. I cannot, however, accede to the propriety of here according to the Vita the standing of a source. Of course, in so doing, Cosenza acted under the spell of the traditional view, to which this article is endeavoring to apply the acid test. He should at all events have taken cognizance of the Servian Life as well, since it is almost as explicit on the subject of the frustration of Vergil's order to destroy the Aeneid as is the Life by Donatus. The Servian Life does not contain the epigram of Sulpicius, but admitting, as does Cosenza, the significance of Ergone supremis as a source of suggestion to Petrarch, there is no compelling reason to send us beyond the poem and the Servian Life in order to account for every- thing that we meet in this context of the Epistle to Vergil. How closely in thought and in motivation Petrarch approaches the poem, will be most apparent if we can bring ourselves to employ for a moment that analytic method to which the pedantic seeker after the well-springs of a poet's fancy often is compelled to resort. Aeneas, says Petrarch, lives in the immortality of the great poem about Aeneas. This is the old motif on which the Roman poets are never tired of ringing the changes; says Tibullus, Quern referent Musae, vivet dum robora tellus, Dum caelum stellas, dum vehet amnis aquas, or, as the author of Ergone supremis has it, Ilium, ilium Aenean nesciret fama perennis, Docta Maroneo caneret nisi pagina versu! (11. 15-16.) Survivor of the burning of the material Troy, Aeneas was, figuratively speaking, facing death in the flames a second time because of Vergil's condemnation of the Aeneid. The hero is saved by the loyal devotion Petbarch's Knowledge op the Life op Vergil 381 of his descendant, Augustus, an act which is a replica of the deed of filial piety to which Anchises owed his escape. Augustus did not respect the wishes of the dying poet, nevertheless the verdict of all time will justify this obduracy. The turn which Petrarch gives to the topos of the second Iliu- persis bears far closer resemblance to the treatment accorded to the theme in Ergone supremis than to the manner in which the motif is handled in either version of the epigram of Sulpicius or the other pieces in which it figures. The fancy that the leading personalities of the Aeneid would be cremated by the flames that should consume the poem is common to Petrarch and Ergone supremis alone among all possible sources. In the latter the theme is driven home by the accumulation of exempla — Creusa, the Sibyl, Dido, and Camilla in the received version, 1 in Recensio /3 the last three — all, except the Sibyl, being figures that had met death in the story, hence would, as it were, feel the flames a second time if the epic were burned. In the line devoted to Dido the paradoxical bis perire is present, what- ever the text we choose. Petrarch adopts this same nuance, but, with a true instinct for effect, heightens the gravity and the pathos of the impending disaster, and emphasizes the service performed for posterity by Augustus, by confining himself to one, and that the most poignant, example of what the destruction of the Aeneid would have spelled, viz., the annihilation in the world of fame after his erstwhile narrow escape from a like fate in the world of myth, of the essential character of the poem. Again, the dominant mood of Ergone supremis is apologetic. Augustus is represented in it as pleading for exculpation because he has set at naught the last mandate of a dying man. The defense put on the lips of the Princeps is that the end justified the act, that he observed a higher law (cf. 11. 18 f.), that such a masterpiece as the Aeneid deserves immortality. Similarly, Petrarch defends Augustus in a vein that seems unmistakably reminiscent of the tenor of what he regarded as the authentic plea of the Princeps. Pietas 1 Whether Petrarch's text of the poem was that found in most manuscripts and in those which constitute our main reliance today, or belonged to the class of the inferior version, /3 in Riese's terminology, I cannot presume to say. The chances would favor the former alternative. In any case our arguments are not affected by the uncertainty. 382 Duane Reed Stuart Augusta reads like a conscious rebuttal of the accusation of bad faith which the imaginary critic is anticipated in the poem as bring- ing. This also is the bearing of the sentence "meritoque supremas contempsisse preces aevo laudabitur omni." In passing, we may note that the verbal resemblance between 11. 27 f. of the better recensio, "iterum sentire ruinas Troia suas, iterum cogetur reddere voces? Ardebit, etc.," and Petrarch's "Arsurum iterum .... secundis flammis" is closer than between the epigram of Sulpicius and Pet- rarch; we may compare Petrarch's supremas preces with supremis verbis 1. 1, suprema voluntas 1. 18. We need not, however, lay great stress on similarities in phraseology; common stock of words tends to accompany common stock of ideas. Furthermore, it must always be borne in mind that in the case of a writer such as Petrarch, whose style so frequently reproduces without conscious effort of memory on his part the locutions of the ancient authors with whom he lived, and who prescribed with reference to the ethics and technique of imitation — "curandum imitatori, ut quod scribit simile, non idem sit" (Famil. 23. 19, Vol. Ill, p. 239 Frac.) — parallelism in diction is bound to be an elusive thing. The analogies that we have noted in spirit and in motifs yield proof, so far as it is possible to effect a demonstration in an attempt to call back the ghosts of a poet's literary processes, that the famous little poem was the source par excellence which suggested to Petrarch the conceit of the second burning, that the context in the Life by Donatus has really no claim to consideration as a collateral source, and, certainly, none as an exclusive source. It may not be amiss, before leaving this topic, to mention an interesting personal application which Petrarch makes of the conceit, although there is nothing in the passage to indicate unequivocally the influence of one source to the exclusion of others. In De con- temptu mundi 3, p. 365 Op., he tells, how, stricken with a serious illness, he had been tempted to consign to the flames with his own hands his unfinished manuscript of the Africa, because he feared that his friends, after the precedent of Vergil's, might play him false. Had he followed this impulse, Africa, always scorched by the rays of the sun, and thrice wrapped in flames by Roman torches, would have suffered at his hands another conflagration! Petrarch's Knowledge of the Life of Vergil 383 II My apology for the riot of destructive criticism to which the previous pages of this article have been so largely devoted must be the inchoate stage at which the discussion of the question has been left by our foremost critics in the field of Petrarchan study. Naturally, those pioneers whose task it is to blaze a trail per una selva oscura may occasionally fail to exercise the meticulous inspection of the route that those who follow in their footsteps find it compara- tively easy to do. By way of supplementing the data thus far presented, I now turn to the consideration of the other passages in Petrarch's works that must challenge the attention of one who is interested in determining the sources from which the humanist derived such knowledge as he possessed of Vergil's life. References by Petrarch to such topics of universal knowledge and widely dis- seminated mention in the testimonia as Vergil's birthplace, his rustic origin, and the like may be dismissed as indices of no moment. In this same class fall allusions to the literary executorship of Varius and Tucca (Senil. 3. 1) to the proposed burning of the Aeneid (Famil. 20. 12, Vol. Ill, p. 40 Frac.) where there are forthcoming no precise indications to declare for one source among the several possible sources. I shall confine myself to a review of passages containing biographical information of such a character as fairly to suggest the possibility of contact between Petrarch and some specific source or sources. In Famil. 10. 4, Vol. II, p. 89 Frac, there is, so far as I have discovered, the sole surface reference in Petrarch to a biography of Vergil. That the Vita to which reference is made is not the Life by Donatus, but the Servian Life, is a fact of prime importance. Not- withstanding its significance for our question, the passage has not been cited in this connection. In elucidating the symbolism of the eclogue entitled Parthenias, Petrarch wrote: "Parthenias ipse est Virgilius, non a me modo fictum nomen; in vita enim eius legimus quod Parthenias, quasi omni vita probatus, dici meruit." Mention of this maidenly surname of Vergil and of the reason of its application to him is found in two of our extant ancient lives of Vergil. (1) In the Suetonian Life, 11. 35-37, we read: "cetera sane vitae et ore et animo tam probum constat, ut Neapoli Parthenias vulgo appellatus 384 Duanb Reed Stuart sit." (2) In the Life by Servius we find a similar comment, evidently derived from the Suetonian Life: adeo autem verecundissimus fuit, ut ex moribus cognomen acceperit; nam dictus est Parthenias; omni vita probatus uno tantum morbo laborabat, etc., Diehl, p. 41. The presence in Petrarch and in the Servian Life of the locution omni vita probatus, by no means the sort of crystallized formula that two writers, independent of each other, would necessarily adopt to express the idea, gives patent proof that Petrarch was here quoting exactly the Servian Life. Noteworthy for us is the obvious fact that he writes as if the Servian Life were for him the canonical biography of Vergil. A close parallel to the passage from the Familiares cited above is to be found in the Epitome which Petrarch composed of his Bucol- ics. Concerning the Parthenias he says: "Istius egloge que prima est in ordine titulus est Parthenias quod nomen interpretatur omni vita probatus. Nam et Virgilius Parthenias vocatus est qui in omni vita probatus in hac egloga primus et principalis pastor intro- ducitur;" see Scritti inediti di Fr. Petrarca pubblicati ed illustrati da Attilio Hortis, Trieste, 1874, p. 359. Here also the echo of the Servian Life is plainly discernible. Let us not be too hasty in formulating on the basis of these passages the conclusion that Petrarch had never read the Life by Donatus. We cannot, however, disguise the trend of the evidence which they offer. In former instances we have seen that when the subject-matter was of a type that might conceivably indicate recourse to the Suetonian Life, we were not forced to the decision that this was the source of which he necessarily or probably availed himself. In the present case, under similar conditions, we can demonstrate his entire independence of the Life by Donatus. A similar deduction may be drawn from a passage in the Itiner- arium, p. 560, Op. Referring to Naples, Petrarch says: "Haec est civitas ubi Virgilius noster liberalibus studiis operam dedit, cum iam patria ilium tua Mediolanum tenerioribus annis discipulum habuisset. Hie se carmen Georgicum scripsisse, hie se ignobili studio fioruisse iucundissime 1 memorat. Hanc dulcem vocat ille Parthenopem .... demum peregre moriens inter extrema suspiria suae meminit 1 The text of the Basle edition is ftoruisseve iucundissime; the Venetian edition of 1501 reads fioruisse verecundissime. Petrarch's Knowledge of the Life of Vergil 385 Neapolis et hue revehi optavit ut quae vivus amaverat vita fructus incoleret." The tradition that included Naples among the seats of Vergil's education is not found in the genuine redaction of the Life by Donatus; herein are mentioned (11. 20-24), sojourns at Cre- mona and at Milan prior to the young poet's departure to Rome. In the Servian Life, on the other hand, we read: "diversis in locis operam litteris dedit; nam et Cremonae et Mediolani et Neapoli studuit." The interpolated Donatus follows the Servian in this detail, inserting after the words found in the genuine version (Diehl, sec. 7), "sed Vergilius a Cremona Mediolanum et inde paulo post," the interpolation: "Neapolim transiit. Ubi cum litteris .... vehementissimam operam dedisset, tandem omni cura .... indulsit medicinae et mathematicis .... se in urbem contulit." We must therefore conclude that it was from the Servian Life that Petrarch derived his information as to Vergil's study at Naples, unless we are prepared to accept as an alternative the view that he used the interpolated redaction of the Life by Donatus. As has been pre- viously pointed out, the present state of our knowledge does not warrant the assumption that the humanistic redaction existed in Petrarch's time. There is, nevertheless, an element in the passage from the Itinerarium which might seem to bear on this problem and hence demands a moment's consideration. I refer to the allusion contained in the last sentence to Vergil's devotion to Naples and to his wish to be buried there. It happens that in respect to this detail none of the extant biographies of Vergil stands so close to Petrarch's statement as does the interpolated Life. In section 54 Diehl, there is an insertion that reads thus: "voluit etiam eius ossa Neapolim transferri, ubi diu et suavissime vixerat; ac extrema valitudine hoc ipse epitaphion fecit distichon : Mantua me genuit, Calabri rapuere, tenet nunc Parthenope: cecini pascua, rura, duces. Translata ergo iussu Augusti eius ossa prout statuerat Neapoli fuere, sepultaque," etc. Here the fact that it was Vergil's last wish that he should be buried at Naples is as explicitly stated as it is by Petrarch. In the corresponding context of the genuine version no direct allusion to an expressed wish is found; we are merely informed 386 Duane Reed Stuabt that Vergil was buried at Naples, and the epitaph is quoted. Pre- cisely these elements and no more are present in the account given by the text of the Servian Life utilized by Petrarch. However, this one instance of parallelism should not lead us to annul the prevalent doctrine, which seems rightly to deny dependence by Petrarch on the interpolated Life. Vergil's fondness for his dulcis Parthenope was, of course, known to Petrarch from the con- cluding lines of the fourth Georgic, partly paraphrased in our passage. The epitaph itself, in which Vergil looks forward to his interment at Naples as fait accompli, implies that the poet entertained the wish and anticipated its fulfilment. But, to account for this context in the Itinerarium, we are not restricted to speculations, plausible though they be, as to what — given a knowledge of the Servian Life — the powers of inference, the imagination, and the vocabulary of Petrarch might have achieved. Jerome, an author whose works were well thumbed by Petrarch, states directly (Euseb. Chron. 1998) what the Servian Life allows to be inferred, that the epitaph was composed by Vergil in his last moments with the expressed anticipa- tion of lying at Naples. Note that Petrarch had merely to embellish the words quern [titulum] ipse moriens dictaverat. 1 The trail of the interpolator doubtless leads back to the Chronicon, unless we care to hazard the suggestion that he had borrowed from Petrarch the element which both have in common. We have a more striking instance of reliance on the works of Jerome, in this case for an incident about which the genuine version of the Life would have been as adequate a source of information. In 11. 186 f. is quoted on the authority of the book of Asconius Pedianus, Contra Obtrectatores Vergilii, Vergil's retort to the critics who accused him of plagiarizing Homer: "sed hoc ipsum crimen sic defendere adsuetum ait: 'cur non illi quoque eadem furta temptarent ? verum intellecturos facilius esse Herculi clavam quam Homero versum subripere'!" This reply was to furnish to future generations of Vergil's partisans their favorite weapon of defense. Petrarch of course employed it. However, in the case of each of the four allusions or reminiscences that his works contain, it is evident that he has not preserved the point of the anecdote as the Life presents it and that 1 For passages indicating knowledge of the Chronicon, see Nolhac, II, 206, n. 3. Petrarch's Knowledge of the Life of Vergil 387 he betrays no acquaintance with that version. Instead, he followed Jerome, who gave the retort a force appreciably different; see Praef. hebr. quaest. in Genesim, Migne, Vol. XXIII, col. 983 : " Hoc idem passus est ab aemulis et Mantuanus vates, ut cum quosdam versus Homeri transtulisset ad verbum, compilator veterum diceretur. Quibus ille respondit, 'magnarum esse virium, clavam Herculi extorquere de manu.'" Now, it will be observed that in the Life Vergil is made rather to emphasize his innocence of the charge, his defense being the sheer impossibility of rifling the treasure house of Homer. Such a procedure is classed among aSvvara, of which Herculi clavam subripere is a typical formula; cf. Macrobius Sat. 5. 3. 16. Jerome gives the story the following color: Borrowing from Homer is cheerfully admitted by Vergil, as an act in itself betokening the possession of superlative poetic power. Examination of the parallel passages in Petrarch will show that he gives only this latter shading: (1) Famil. 22. 2, Vol. Ill, p. 126 Frac: "etsi enim non me lateat quosdam veterum Virgiliumque ante alios versus innum- eros non modo e graeco in latinum versos, ubi abstulisse clavam Herculi gloriatur." (2) Famil. 24. 12, Vol. Ill, p. 298 Frac: " quae [de Virgilii imitatione] tamen ex ordine ipsis in Saturnalibus scripta sunt, quamvis hoc loco ille suus iocus innotuerit, cui cum obiiceretur ab aemulis quod versus tibi [i.e., Homer] tuos eriperet, magnarum virium esse respondit auferre clavam Herculi;" as Cosenza (op. cit., p. 197) has pointed out, Petrarch confused in memory the contents of the passages in Jerome and Macr. Sat. 5. 3. 16 and made a slip in quotation. (3) Res memor. 2. 3, p. 419 Op.: "nee illepide Virgilius, cum sibi exprobratum esset quod versus Homericos abstulisset et in operis sui congeriem redegisset, respondisse traditur magnarum esse virium auferre clavam de manu Herculis, non dissimulans se illis versibus non praecario sed pro suis uti." (4) Variae 62, Vol. Ill, p. 476 Frac: "solus adversum agmina clavam de manibus Herculis extorsisti." This is evidently a reminiscence of the passage in Jerome; cf. extorsisti and extorquere. The treatment accorded by the Life to the rubric of Vergil's detractors is conspicuously detailed; see 11. 176-92. In addition to the passages just cited, Petrarch alludes to this topic in the fol- lowing passages: (1) Famil. 1. 1, Vol. I, p. 31 Frac: "quid Virgilio 388 Dtjane Reed Stuart maius habuit lingua latina ? Reperit tamen ille, qui non poetam sed raptorem alienarum inventionum et translatorem diceret. Ipse autem et ingenii fiducia et iudice fretus Augusto, alto animo invi- dorum verba despexit." (2) Famil. 4. 7, Vol. I, p. 217 Frac: "ipsorum quoque Maronis ac Flacci aetas non aequa tantis ingeniis fuit, quorum alter divini spiritus poeta dum vixit, aemulorum bellis sine fine vexatus, ut alienorum operum deflorator carpitur." In these references to the disparaging criticism directed against Vergil by his contemporaries we can detect primarily the influence of the passage in Jerome. Certain it is that there are no features of content or language that suggest the influence of the Suetonian Life. In contrast with the too copious data found in the ancient com- mentators and biographers on the subject of the eviction of Vergil from his farm, Petrarch shows no great interest in this feature of the biographical tradition. In Famil. 11. 5, Vol. II, p. 115 Frac, he writes: "Virgilio Augustus Caesar agrum reddiderat sed quern ipse praeripuerat." This bare epitome of the traditional account, based on exegesis of the first Eclogue and set forth in detail by Servius in the prooemium to his commentary on the Bucolics (Thilo-Hagen, III, 3) offers little that is instructive for our purpose. However, it may be noted that in the Life by Donatus credit for protecting the poet and for active intercession in his behalf is given to Pollio, Alfenus Varus, Cornelius Gallus, and Maecenas. No mention is made of the clemency of Augustus, though naturally he was the final court of appeal. By Servius, just as by Petrarch, the Princeps is explicitly made Vergil's benefactor; cf. Servius, cit.: "postea ab Augusto missis tribus viris et ipsi integer ager est redditus, etc." What Petrarch tells us about the material fortunes in general of Vergil bears no mark suggesting that the Life, with its specific inventory of the poet's property (11. 40-43), was a source of knowl- edge. In De remediis utriusque fortunae 2. 9, to a list of poor but illustrious Roman men of letters, including Plautus, Horace, Pacu- vius, and Statius among poets, Vergil is added: "Inops demum aliquando Virgilius donee praeter suum morem opes ingenio acces- sere." In Senil. 2. 2 we find: "nunquid ergo aut Virgilius multo auro ditatus a Caesare fuit obscurior quam dum rure primo depulsus exul atque inops Romam peteret?" Here also the specific allusion Petrarch's Knowledge op the Life of Vergil 389 to Augustus as Vergil's benefactor points to the first Eclogue, to Servius on Aeneid 6. 861, as we shall see in a moment, a locus classicus for Petrarch in connection with another event in Vergil's life and which records the liberal honorarium received for the sixth book of the Aeneid as follows: "qui [i.e., Vergil] pro hoc aere gravi donatus est, id est massis." There was also Horace Ep. 2. 1. 246 testifying to the bounty of Augustus: At neque dedecorant tua de se iudicia atque munera, quae multa dantis cum laude tulerunt, dilecti tibi Vergilius Variusque poetae. Another instance plainly betokening ignorance or neglect of the Life by Donatus is furnished by Petrarch's version of the incidents attendant on the famous recitatio of the sixth book of the Aeneid. In the Life, 11. 109-12, we have the dramatic story of the swoon of Octavia at the words tu Marcellus eris; nothing is said as to the emotions aroused in the breast of Augustus. Servius on Aeneid 6. 861 paints the scene differently: "et constat nunc librum tanta pronuntiatione Augusto et Octaviae esse recitatum, ut fletu nimio imperarent silentium, nisi Vergilius finem esse dixisset"; there follows the allusion to the reward as transcribed above. It was evidently to the passage in Servius that Petrarch reverted when penning the following words in De remediis 1. 114: "Marcel- linus .... quantum putas exspectaretur ab avunculo, qui eum sic dilexit, ut Virgilianum illud carmen nobilissimum .... sine lacrimis audire non posset et actori silentium imperaret," although, as the reader can discern, he slightly exaggerates the reaction of Augustus. He might, of course, have utilized merely the testimony of this passage as to the poignancy of Octavia's grief for her child. Instead, he preferred to dignify this topic and at the same time to enhance the effect of his exempla by devoting to it a separate allusion based upon a different source. He continues: "quomodo mater Octavia quae ilium sic amavit, ut usque ad ulti- mum vitae suae finem, quasi tunc extinctum, continue deploraret, omnemque consolatorem, non tantum contemneret sed odisset?" In Variae 25, p. 389 Frac, and in Senil. 10. 4, p. 876 Op., appears similar allusion to the life-long persistency of Octavia's mourning 390 Duane Reed Stuart and to the obduracy with which she refused to be comforted. As comparison will show, underlying these three passages is the same source, viz., Seneca, Ad Mar dam de consolatione 2. 4: "Nullum finem per omne vitae suae tempus flendi gemendique fecit," etc. Without indulging in the luxury of an argumentum ex silentio, I may at least venture to say that it is astonishing that the spectacular appeal and the effectiveness of the anecdote as related in the Life, if Petrarch had ever read.it there, did not force it on his memory in connection with some one of these contexts. In Epist. metr. 1. 2, Vol. Ill, p. 90, col. 2 Op., Petrarch alludes to Vergil's laborious methods in composition : O utinam nostra quondam tarn larga Maroni Copia dicendi! numquam, mihi crede, laborem Lentus inexpletum seros traxisset in annos. The time spent by Vergil on his works, the merciless pruning to which at times he subjected them, the deliberation and the capricious- ness that in general characterized his literary modus operandi, are set forth in the Life by Donatus, 11. 78-90, with considerable detail. In our passage, however, the poet's lack of fluency seems to be an inference resting merely on the fact that the composition of the Aeneid was so long protracted. The Servian Life furnishes sufficient testimony that the poem was left without the final revision. This is the only topic connected with Vergil's literary technique at which Petrarch and the Life by Donatus come in contact. He thrice alludes to Vergil's lack of felicity as a writer of prose, viz., in Famil. 4. 15, Vol. I, p. 238 Frac; Res memor. 2. 2, p. 410 Op. (twice). The source of this assertion is Seneca, Contro. 3, p. 243 K, to whom surface reference is made in each of the three instances. A chapter in the De vita solitaria treats of the ancient poets who preferred to live a life apart from "the madding crowd's ignoble strife." Among the exempla Vergil of course finds a place: "Quid Virgilium nostrum loquar ? qui f ugiens urbem Romam ubi et ingenii laude et toto orbe regnantis principis amicitia florebat ac solitariam libertatem petens, mortem quidem immaturam obviam habuit, quae omnibus eum talibus cutis absolveret. Ille sic censebat ope solitudinis opus esse ut divinum ilium suum opus posset absolvi" (De vita sol. 2. 7. 2, p. 279 Op.). Petbabch's Knowledge op the Life of Vergil 391 The Life by Donatus recites with greater circumstance than any of our other extant sources Vergil's antipathy toward life at Rome and his preference for a secluded existence in the "provinces," see 1. 37: "Romae, quo rarissime commeabat;" 11. 41-44: "habuit .... domum Romae Esquiliis .... quamquam secessu Cam- paniae Siciliaeque plurimum uteretur;" 11. 123-25: "anno aetatis quinquagesimo secundo impositurus Aeneidi summam manum statuit in Graeciam et in Asiam secedere triennioque continuo nihil amplius quam emendare." In this case, at first sight, Petrarch seems more nearly to coincide with the Vita than in any of the other instances that I have added to those utilized by Sab- badini. Especially is this true if we are to see in the sentence "ille .... absolvi" allusion to Vergil's departure to the East, there to put the finishing touches to the Aeneid. The Life by Donatus, alone of our extant sources, expresses thus concretely the intention with which Vergil undertook the journey. In Petrarch's exemplar of the Servian Life nothing is said of that desire to revise the Aeneid which impelled Vergil to seek new scenes; indeed, the poet's presence in Apulia at the time of his death is quite unmotivated. Nevertheless, before this passage in Petrarch can be classified as evidence corroborative of the use of the Suetonian Life by him, it is necessary to assure ourselves that what he tells us here must needs point inevitably to that source only. I cannot regard the data as thus definitive. Petrarch did not need the Life by Donatus to inform him that Vergil lived by choice away from Rome. Tradition alone connected the major portion of the poet's life indissolubly with Naples. There were also the concluding lines of the Georgics, the devotion to a country life pervading both Eclogues and Georgics, and voiced as an object of personal longing in Georg. 2. 483 f. Petrarch knew from the Servian Life that Vergil died while traveling in Southern Italy. Analysis of the passage in De vita solitaria shows that the sentence "ille sic," etc., looks back to the preceding sentence. Here the members fugiens urbem Romam and solitariam libertatem petens certainly do not justify the assumption that Petrarch could be referring onlyto the departure from Italy chronicled by the Suetoni- an Life. His language is not thus restrictive. He had in mind merely Vergil's chronic dislike of life at the capital, his habitual taste 392 Duane Reed Stuart for seclusion, manifested by his fondness for Naples and his residence there, his death while he was visiting places of a bygone greatness, in the idyllic regions of Southern Italy. Such an expression as "in quest of the freedom of solitude" is better applicable to a sojourn in the reposeful country " qua niger umectat flaventia culta Galaesus " than to a journey to the East. The reason assigned for Vergil's retirement — the desire to finish his great epic — is that which above all others would naturally have suggested itself to Petrarch in this connection, intent, as he was, on lauding the advantages of the life lived by the literary recluse, knowing as he did that, in spite of the long years spent in travail apart from "the seats of the mighty," the Aeneid was left unfinished. An item in which there is approximate agreement between Petrarch and the Suetonian Life is the age of Vergil at his death. Compare with 11. 122-25 cited above this passage in Semi. 16. 2: "septuaginta [annos] vixit Ennius, totidem Horatius Flaccus; duos et quinquaginta Virgilius nostra etiam aetate breve tempus." As a matter of fact, Vergil died just before the completion of his fifty-first year — he would have been fifty-one had he lived to the ides of October, 19 b.c. The Life attributed to Probus is that one of the ancient sources that is most precise in this respect : " decessit in Calabria annum agens quinquagesimum et primum." The loose calculation of the Suetonian Life may be the result of carelessness. However, a more plausible supposition is that Vergil, when he started on the journey that was to be his last, was so close to the beginning of his fifty-second year — in modern parlance was so near "fifty-one and going on fifty-two" — that the author of the Vita felt justified in disregarding the few weeks intervening. Perhaps, as is likely to be the case in numerical approximations, a round even number had an appeal. Petrarch's "fifty-two years" of course transcends in license even the statement of the Vita, yet might conceivably have been based upon it. The question is whether this source furnished the only likely authority for the number adopted by Petrarch. As a possible alternative, we should think first of all of Jerome's version of the Chronicon of Eusebius. This work merely sets down the dates of Vergil's birth and death as they are given in the Suetonian Life, undoubtedly the source from which Jerome derived them. He Petbabch's Knowledge of the Life of Vergil 393 did not record the age of Vergil at his death, although he followed this practice elsewhere, in the case of Horace, for example (Migne, p. 440) and Ennius (Migne, pp. 417-18). An exact computation based on the Chronicon would not yield the number fifty-two for Vergil's mortal span. Furthermore, the numbers that figure in this passage were not the products of painstaking calculation or of the utilization of annalistic data. Petrarch was clearly drawing on the stores of memory; nor did he after the approved fashion of modern scholar- ship trouble himself to turn pages in a deliberate effort to verify his recollection. Hence his error of thirteen years 1 in respect to the duration of Horace's life, although the number is given correctly both in the Suetonian Life of Horace and in the passage of the Chronicon referred to above. Have we then in this instance an argument undeniably impressive in favor of the view that Petrarch knew and used the Life by Don- atus ? Without for the time being appealing to the rather consistent trend away from such a conclusion maintained by the evidence, new and old, which we have thus far examined, we can cite one consideration which, by itself, makes against the absolute cogency of this parallel. The authority of the Suetonian Life had centuries before Petrarch's time conferred a traditional standing on the number fifty-two as the sum-total of the years of Vergil's life. This is shown by the fact that the number was adopted in the imaginary epitaphs of Vergil in the composition of which the poetasters of later years were fond of trying their skill: cf. Anthol. Lat., Baehrens IV, p. 129, 11. 21-24; p. 130, 11. 45-48 = Riese I, 2, Nos. 560, 566; I quote the pertinent distichs: Hie decern lustris geminos postquam addidit annos Concessit fatis et situs hoc tumulo est; Iamque ad lustra decern Titan aecesserat alter, Cum tibi me rapuit, Mantua, Parthenope. The authors of the epigrams permitted themselves the same liberty of computation that we have noted in the case of Petrarch. There seem to be reasonable grounds for doubting that Petrarch's statement 1 The sessanta of Fracassetti's Italian translation, II, 535, is, I surmise, a slip or a typographical error, not a deliberate correction. The same number cannot be right for both Ennius and Horace, in any case. 394 Dtjane Reed Stuart was inspired by memory of the sentence in the Life — memory, I say, because the passage as a whole does not reveal the results of research and verification. It may well be that he merely set down for the years of Vergil's life the approximation which, established by tradi- tion, had become a commonplace of information with him as well as with the world of learning. We have now reviewed the passages in Petrarch that furnish points of departure for an attempt to discover whether, as has been universally held, Petrarch knew the genuine version of the Suetonian Life, and, if so, to what extent he derived from it such information about the career and the personality of Vergil as he gives us. Con- cerning the latter question there can be, in the light of the evidence that we have marshaled, no two opinions. In the great majority of the instances that have been examined we have seen that Petrarch turned to the Servian Life, to the Commentary of Servius, and to his well-read authors for data which the Life by Donatus was equally, and, in some cases, better fitted to supply. Among possible biograph- ical sources the authority of the Servian Life is pre-eminent. It is apparent that, if Petrarch had read the Suetonian Life, his reliance on it was not even semi-occasional. I am not without hope that the reader of judicial mind has been persuaded to subscribe to my belief that, in the present state of our knowledge, we are not warranted in assuming that Petrarch used the Life by Donatus. At all events, we may justly contend that the current theory has been revealed as resting on supports more tenuous than has hitherto been realized, that such assertions as Nolhac's "il lisait, en effet, la vie de Virgile attribute a Donat" and Sabbadini's "il Petrarca adoperava certamente .... la biografia breve di Donato" would better have been couched in more conservative phraseology. A candid estimate, it is safe to say, would rate highest as indications of acquaintance on the part of Petrarch with the Suetonian Life his allusions to the age of Vergil at death, to the letters written by Augustus to Vergil, and, lastly, his knowledge of the couplet of Propertius. In each of these three instances, examined by themselves, we discovered, unless I am oversanguine, reasonable grounds for denying their absolute cogency; the cumulative effect of the evidence as a whole makes potently Petrarch's Knowledge of the Life of Vergil 395 against stressing unduly the testimony that they might be supposed to offer in favor of the traditional view. Of great significance, to my mind, is the fact that, as we have seen, the Servian Life is the only biography of Vergil to which Petrarch directly refers. At a relatively late period in his life he implies that for him the Servian Vita was the biography of Vergil. The date of the letter in which the passage in question occurs is, as fixed by Fracassetti, 1348. However, the passage may fairly be regarded as symptomatic of the extent of his knowledge of Vergilian biographies at a later epoch in his life, that is, the five or six years subsequent to 1359 during which he was re-reading and editing his letters for publication. We know that retention of language and subject- matter precisely as they had existed in the original drafts was not a part of his program; he did not hesitate to introduce changes and to make insertions representative of the stages of knowledge and taste to which he had attained in these later years of his life. 1 I must not neglect to mention certain facts of an external nature which form an element in our discussion. The meritorious researches of Sabbadini 2 have demonstrated that the Life by Donatus in its genuine version was known to friends and contemporaries of Pet- rarch. Benvenuto Rambaldi, surnamed da Imola, to whom Senil. 15. 11 is addressed, derived from 11. 49-50 the authority for the assertion, 1. 43, "Virgilius fuerat tardissimus in sermone . . . . ut scribit Donatus super Virgilium." The surface reference is a notable exception to the ignorance of the authorship of the Vita generally prevalent in the fourteenth century. The works of Boccaccio also seem to reveal acquaintance with the Life 3 and he apparently pos- sessed a copy of it, although he did not know it under the name Donatus. These facts do not by any means enhance the likelihood that Petrarch had perused the biography, even though we grant that he may have known of it by hearsay. It is almost superfluous to point out that the presence of a given monument of ancient literature in 1 Cf . Sabbadini, "II primo nucleo della biblioteca del Petrarca," Rendic. del r. ist. Lombardo, XXXXIX (1906), 369. * "Le biografie di Virgilio antiche medievali umanistiche," Studi Hal. di filol. class., XVI (1907), 242. « Ibid., pp. 243-14. 396 Duane Reed Stuart the collections of Petrarch's friends is no proof that he had ever read the work in question. As we know, in the library of Boccaccio were several ancient authors of the first rank, Tacitus and Martial 1 among them, whose works remained closed books to Petrarch, not- withstanding the friendship and the community of intellectual interests that held good between these two leaders of humanism. Boccaccio's acquaintance with the Appendix Vergiliana was quan- titatively greater than Petrarch's. Hence it should not cause surprise that, even though his coenthusiast in the study of antiquity had a manuscript containing a biography of Vergil more elaborate than the Servian Life, Petrarch should not have had this longer and more detailed biography by him and should not even have read it. The accessibility of the Servian Vita and the possibility of supple- menting it by the copious material available to him in the works of many of the authors who were the constant companions of his waking hours, may well have tended to satisfy him with what these sources, endowed as they were with the prestige of great names, could teach him about Vergil. We must also bear in mind that in Petrarch's lifetime the authoritative name of Donatus had not been definitely and universally attached to the biography, that in general it remained anonymous until toward the close of the fourteenth century. 2 Therefore, external data of this character cannot vitiate the conclusions that we have based on internal evidence. Indeed, a set of facts which I have reserved for final consideration comes tantalizingly near to offering welcome corroboration of my belief in Petrarch's ignorance of the Life by Donatus. As Sabbadini 3 has reported, Francesco Nelli, the Simonides who figures so frequently as the recipient of letters from Petrarch, in a letter written to Petrarch in the year 1351 quotes from the Life, 11. 48-49 as follows: "scis quoque, ne iocundum hoc obmittamus, Virgilium semel causam egisse." 4 Now, on the margin of the manu- script there is the following note, written, as Cochin testifies, in a 1 Sabbadini, Le scoperte dei codici latini e greci ne' secoli XIV e XV, pp. 28 f. 2 Of the manuscripts listed in Brummer's apparatus, which are anterior to the fifteenth century and contain the Life in full, one only (Parisinus lat. 11308) is by caption connected with Donatus: cf. Sabbadini, Studi Ital. di filol. class. XV, 235. " Studi Ital. di filol. class. XV, 242. * H. Cochin, Lettres de Francesco Nelli a Petrarque (Paris, 1892), p. 182. Petrarch's Knowledge of the Life of Vergil 397 hand different from that which appears in the rest of the manuscript: "Non memini me legisse de Virgilio, nisi forte iste vir intelligat Virgilium egisse causam pro possessionibus sibi ablatis et postea restitutis." Apropos of this note the editor comments : " Je n'oserais pas afnrmer que cette note soit de Petrarque. " It is tempting, indeed, in the light of our results, to claim the note definitely as the product of Petrarch's pen. However, in the face of Cochin's hesitation and until inspection of the note and its comparison with other specimens of Petrarch's handwriting can be undertaken, dogmatism would be ill-timed. We must rest content with suspicions and with the hope that these may be verified. If I may be permitted to lapse into a personal vein, I should like to say that my conclusions with reference to Petrarch's relation to the Life were formed quite without recourse to the argument suggested by the note. 1 Knowledge of it came merely as agreeable, though not unquestionable, support of a position already assumed. This is certain: The — let me say — scholiast's ignorance of the provenience of the statement in Nelli's letter and his suggestion that it is an inference derived from the first and the ninth Eclogues savors strongly of what we have discovered elsewhere as to Petrarch's methods of exegesis, and conforms pre- cisely with the conclusions that we have framed on the basis of the internal evidence. In the works of Petrarch there are a few other contexts which, although they do not offer evidence pertinent to the issue that has primarily inspired this paper, nevertheless present such interesting problems in respect to the sources of Petrarch's knowledge of the life of Vergil as properly to claim consideration here. The assertions made in these passages are either at variance with the information furnished by the sources on which we have seen Petrarch chiefly relied, or even overstep their testimony. We are thus confronted with the question whether he had access to sources not comprised in our modern apparatus. In this hypothesis Sabbadini took refuge in order to account for a remark made by Petrarch in Fatnil. 6. 3, Vol. I, p. 324 Frac, to the effect that Vergil became prematurely gray. Such a physical 1 Cochin's book was not easily accessible to me. I owe to the kindness of my friend Professor N. G. McCrea a transcript of the essential context. 398 Duane Reed Stuart characteristic of Vergil is not specifically mentioned by any ancient source. The passage is: "et Numa Pompilius, cuius supra menti- onem feci, prima aetate canus fuit, et Virgilius poeta." Besides this passage cited by Sabbadini there are two others in which Petrarch portrays Vergil as a "good gray poet." (1) In Senil. 8. 1 he adds to the exempla of Numa and Vergil that of Domitian: .... "Virgilii iuvenis barba candidior et Domitiani adolescentis coma senescens." Here barba candidior, an echo of Eclogues 1. 28, gives a plain hint as to the source of Petrarch's notion. (2) In De con- temptu mundi, dial. 3, p. 803, ed. of 1649, the riddle is solved by Petrarch himself. Commenting on his own gray hairs, he consoles himself by recalling the prematurely silvered locks of Domitiam and Numa, then continues: "Nee poeticum defuit exemplum siquidem Virgilius noster in Bucolicis, quae 32 aetatis anno scripsisse eum constat, sub persona pastoris, de se ipso loquens ait: Candidior postquam tondenti barba cadebat." Thus we see that it was merely the allegorical identification of Tityrus and Vergil, a theory which ages of criticism more modern than Petrarch's have been slow to relinguish, that lay at the root of his assertion. In this connection it is interesting to see that Petrarch flatly dissents from the note of Servius on this line of the Eclogue. The commentator rejects for this context the supposition that Tityrus equals Vergil, hence suggests " aut mutatio personae est ut quendam rusticum accipiamus loquentem, non Virgilium per allegoriam" or, as an alternative, proposes to couple candidior with libertas. Thus, to find Petrarch's informant in this instance we are not led beyond the sources that lie within our ken. The assignment of the composition of the Eclogues to Vergil's thirty-second year is a surprising departure from the testimony of Servius as given in the prooemium to the commentary on the Eclogues, Thilo-Hagen III, 3, in the note on Eclog. 1. 28, and, a third time, in the note on Georg. 4. 564. In all these places as well as in the Life attributed to Probus, we are told that Vergil was twenty- eight years of age when he wrote — meaning in lingua scholastica — began to write the Eclogues. Our other data on the chronology of the Eclogues are limited to what can be learned from the poems themselves and to the statement found in both the Life by Donatus Petrarch's Knowledge of the Life of Vergil 399 and the Servian Vita, that the composition of the Eclogues covered a period of three years. Owing to the lamentable inaccuracies which disfigure the texts of the greater part of the works of Petrarch, we cannot be certain that we have not to do with a blunder founded on a misread Roman numeral 1 and perpetuated in our editions. On the other hand, if thirty-two be the number that came from the pen of Petrarch, we can but assume another lapse of memory in recalling figures, and a failure to verify. In none of the three passages in Servius is any trace of a variant indicated by Thilo-Hagen. If Petrarch had said that Vergil finished writing the Eclogues in his thirty-second year, we could conjecture that the computation represented the sum of the twenty-eight years appearing in Servius and the triennium recorded in the Servian Life. However, we cannot read this into Petrarch's words, scripsisse eum constat, any more than we can into the statement of Servius (Thilo-Hagen, p. 3, 1. 26): "sane sciendum Vergilium XXVIII annorum scripsisse bucolica." Furthermore, such a deliberate modification of the number in Servius as this suggestion would presuppose is most unlikely for the reason that it is the smaller number that would by preference commend itself to Petrarch here; the earlier the period in Vergil's life to which the premature grayness can be assigned, the more effective the exemplum for Petrarch's purpose. Again, for Petrarch, unversed as he was in the intricacies of the data on which depends the determination of the relative chronology of the Eclogues, the first Eclogue would undoubtedly be the earliest in point of time. It was still so regarded by Charles de la Rue in the latter part of the seventeenth century, pioneer though he was in the field of Vergilian chronology. There is another piece of biographical data, the provenience of which is by no means obvious. In De remediis utriusque fortunae, 1. dial. 80, entitled "De excellente praeceptore," Ratio, one of the interlocutors, contends that gifted teachers do not invariably produce great pupils, that genius may attain fruition apart from, or in spite of, formal instruction received from a master. Among the exempla are numbered Vergil and Horace: "Nullum Virgilio praeceptorem 1 In XXVIII, VI might as the result of a blot or of some analogous mishap easily coalesce into a third X. 400 Duane Reed Stuaht legimus. Horatius Flaccus de suo nihil, nisi quod plagosum dixit, verberum puerilium, credo, meminerat." This assertion of Petrarch as to Vergil, the interpretation of which I shall in a moment discuss, has, so far as I have discovered, awakened but a single echo in subsequent biographical criticism. Sebastian Corrado, a professor at Bologna, who died in 1556 and who wrote a biography of Vergil marked, ut temporibus Mis, by critical acumen and independence of judgment, devoted a few words to a refutation of Petrarch's statement. This biography is to be found in Henri Etienne's editions of Vergil, issued in 1583 and 1599. Its repute lasted until the seventeenth century, as its inclusion in Taubmann's edition, Wittenberg, 1618, testifies. Corrado writes: "nullum praeceptorem Virgilio legimus (imquit Franciscus Petrarca) male: nam Macrobius autor est eum Parthenio grammatico in litteris Graecis usum fuisse: Servius Sironem eius doctorem in rebus Epi- cureis agnoscit." Corrado's exceptions are well taken; Servius on Eclog. 6. 13 and on Aen. 6. 264 is authority for the statement that Vergil studied the tenets of Epicureanism under Siro. The passage in Macrobius is Sat. 5. 17. 18: "versus est Parthenii quo grammatico in Graecis Vergilius usus est." The clause quo .... est, suspected by Ian and bracketed by Eyssenhardt, appears in all the manuscripts of Macrobius except P, which shows an erasure. It is practically certain, therefore, that Petrarch had read the clause in his exemplar, although this has not, so far as I know, been identified. It is puzzling to explain how Petrarch came to make an assertion so specifically contradicted by the testimony of authors to whose works he had such constant recourse and knew so thoroughly as he did Macrobius and Servius. Was it again simply a case of lapsus memoriae'! The other alternative is that he chose deliberately to neglect evidence which he had encountered, if not in two sources with which he was intimately acquainted, then certainly in one. Intent though he was on making out a case for the heaven-inspired, self-taught genius, we can scarcely believe that he would have resorted so far to ex parte pleading as gratuitously to suppress evidence conflicting with his thesis. Nothing that we know concerning the spirit and the method of his scholarship justifies so harsh an accusa- tion. If he consciously refrained from taking into account discrepant Petrakch's Knowledge of the Life of Vergil 401 testimony, it would rather be that he did not set a high value on it, a supposition which does not accord with his regard for Macrobius and Servius. If we may judge from the form in which Corrado cast his com- ment, he understood Petrarch to mean that Vergil had never had a teacher, and that the assertion was based upon a categorical state- ment to that effect found in some ancient author whose identity Was a mystery to the later critic. For he goes on to say: "Nee video quern secutus autorem Franciscus idem Petrarca scripserit Marcel- lum, Octaviae filium, a Virgilio reprehensum vel irrisum fuisse quod avibus delectaretur." This reference to a fling of Vergil's at a hobby of the young Marcellus is found in De remediis 1. dial. 32, p. 106, ed. of 1649. This passage will claim further attention. Corrado's interpretation, or, as I believe, misinterpretation of the sentence "nullum Virgilio praeceptorem legimus," is also indicated by his transposition of the dative to a place following praeceptorem. No source known to us contains any such statement about Vergil as Corrado assumed Petrarch to make here. What we have learned in a recent instance recommends the wisdom of seeking first some other explanation than the hypothesis that Petrarch used biographical sources now lost to us. Let us see if a plausible solution may not he in a correct understanding of what Petrarch says. Corrado, I believe, failed to grasp the precise shade of meaning that Petrarch intended to convey by the sentence "nullum Virgilio praeceptorem legimus." The literal sense is "we read of no teacher for Vergil," the dative, as, indeed, its intermediate position would show, being dependent on praeceptorem. This construction, though more in the form of Sallust, Livy, and Tacitus than that of Cicero, needs no justification, of course. This use of lego has, we recall, ample precedent in good authors; cf. Cicero, De imp. Pomp. 28; Pro Marc. 28; Cato Mai. 20; Pliny Ep. 7. 19. 7; it is strictly in conformity with the Latinity of Petrarch; see, for example, Variae 32, pp. 389-91 Frac. By interpreting in this way the passage before us we obtain, in my opinion, the most plausible explanation of the genesis of Petrarch's statement. I should assume that he had in mind at the moment of writing, to the temporary exclusion of his other sources, the data 402 Duane Reed Stuart offered by the Servian Life on the subject of Vergil's education. We read there "diversis in locis operam litteris dedit," a statement which certainly to Petrarch would have implied that Vergil had undergone a formal training. The name of a teacher is not given. It is easy to see, therefore, how Petrarch could write " we read of no teacher of Vergil," in the Vita of the poet, that is, where such an allusion might be expected to occur. It certainly does less violence to the probabilities to surmise that under the spell of his familiar, formal biographical source he forgot contradictory evidence existing in isolated contexts elsewhere, than to hold that he here gave vent to the positive declaration that Vergil had no teacher. Against accepting this alternative view is the fact that this would be the sole instance among all his allusions to Vergil in which a clue to his source resides neither in the traditions nor the literary testimonials known to us. Lastly, exalted as was Petrarch's estimate of Vergil's genius, I find it difficult to entertain the belief that he was capable of regarding the poet as entirely self-made. Such a notion on the part of Petrarch would not square with what he had read in the Servian Life about Vergil's sojourn and study at the academic center of Milan. Thus far in our enumeration of Petrarch's references to the life and the personality of Vergil, we have not found it necessary to shift our gaze from the figure of the poet as it is sketched in our literary sources. Petrarch's Vergil has been the Vergil of Servius, Macrobius, Jerome — in fine, to adopt the caption of Comparetti, the Vergil of literary tradition. We know in general that Petrarch's conception of Vergil was uncontaminated by the legendary balderdash and the old wives' tales which had attached themselves at Naples to the figure of Vergil. For Petrarch, as his biographers have pointed out, the Mantuan was neither charlatan nor thaumaturge. Petrarch's oft-quoted rejoinder to King Robert of Naples, "nusquam me legisse magicarium fuisse Virgilium," is a classic indication of his attitude. One passage, however, seems to lead us away from the beaten paths of the literary tradition. I refer to the story cited by Corrado from De remediis 1. 32, to the effect that Vergil scoffed at Marcellus' devotion to birds, or rather, as the context and the title of the dialogue, "De venatu et aucupio" indicate, to falconry. The Petrarch's Knowledge of the Life of Vebgil 403 passage reads thus: "atqui ducum et illustrium multos equis, quosdam canibus delectari solitos audivimus .... avibus nullum fere : unde et irrisum f erunt a Virgilio Marcellinum, Augusti nepotem, quod dare his operam adolescentulus videretur." This anecdote, the unique significance of which has been over- looked by students of Petrarch, belongs to a lower stratum than any of the humanist's other allusions to Vergil's career. Corrado, it will be recalled, failed to discover any authority for the assertion. Modern scholarship is not driven to a like confession of inability, for a passage in John of Salisbury, though perhaps not itself the direct source of Petrarch's knowledge of the story, at least paves the way to conclusions. This passage, which both Comparetti 1 and Schaarschmidt 2 have utilized, is our locus classicus of information concerning the famous fly which Vergil is said to have constructed of bronze and set up in Naples as a talisman to rid the city of a plague of flies. See Polycraticus 1. 4: "fertur vates Mantuanus interrogasse Marcellum, cum depopulation! avium vehementius operam daret, an avem mallet instrui in capturam avium, an muscam informari in exterminationem muscarum. Cum vero quaestionem ad avunculum retulisset Augustum, consilio eius praeelegit ut fieret musca, quae ab Neapoli muscas abigeret et civitatem a peste insan- abili liberaret." This story John undoubtedly had heard during one of his visits to Italy; see Schaarschmidt, p. 31. As Comparetti has shown, we have here a tale belonging to the cycle of Neapolitan folk-stories told about Vergil, the great local celebrity. Among other vagaries, the popular legend made Marcellus the duke of Naples and Vergil his prime minister. The supposition that Petrarch had read the works of John of Salisbury would seem, on a priori grounds, not improbable. Gug- lielmo da Pastrengo, a correspondent of Petrarch's, cites the Poly- craticus four times, though without naming the author. Sabbadini, 3 on the basis of two indications mentioned by him, assumes Petrarch's indebtedness to the treatise. With the exception of this passage, I am able to contribute no striking case of parallelism, and in 1 Virgilio nel medio evo, 2d ed. (Florence, 1896), II, 36. 1 Johannes Saresberiensis (Leipzig, 1862) , p. 98, n. 2. 'Rend, del r. ist. Lomb. di sc. e litt., XXXIX (1906), 387. 404 Duane Reed Stuart this present instance Petrarch's version of the story seems rather to separate him from the Polycraticus. I should not lay stress solely on the absence of allusion in Petrarch to the talismanic fly. This detail, as smacking of practice of the black art and foreign to his own rationalistic conception of Vergil, he might have been moved to banish from his account, even if the passage in the Polycraticus had been the source of his knowledge of the story. But we note that, according to Petrarch, the central point of the incident is Vergil's derision of Marcellus for indulgence in fowling because this was a bizarre pastime, unsanctioned by the approval of the sporting world. This element cannot be read into the account of John of Salisbury. It is simply a case of an option offered in good faith between two gifts, one subserving the interests of the city, the other destined to further the selfish pleasure of Marcellus alone. If John of Salisbury were Petrarch's informant, the original was drastically adapted to point a different moral and to adorn a dif- ferent tale, or a hazy memory of the version in his source led him to give the anecdote a different turn. The verbal resemblances that are discernible inhere in common locutions. On the whole, it seems more likely that Petrarch gleaned the story from another source in which the incident was cast in the form in which he has handed it down. We cannot say whether, as John before him, he had heard the story at Naples where tales of this sort were rife, or whether he had come upon it in some wonderbook of the type of the Gesta Romanorum or the Cronica di Partenope, such as were written to chronicle the deeds of "Master Vergil" in the thirteenth and fourteenth centuries. The significant thing, after all, is that Petrarch in this one instance, but in this one only, did not hold aloof from a story emanating from a domain apart from the classic literary tradition which he ordinarily followed in touching on topics connected with the biography of Vergil. Princeton University