existing MS. is older than the 14th century (Teuffel, 211, 9). Of some 250 MSS. of Horace, the oldest are of the 9th century: of Tibullus, none earlier than the 13th; of Ovid, Caesar, Seneca, Juvenal, Persius, and Q. Curtius, none earlier than the 9th. The fact that of so many Latin authors (of whom those here mentioned are but a sample) good, and in most cases the oldest, MSS. date from this period would of itself suggest the inference which history confirms, that it was a period of marked revival in learning and consequently in the production or reproduction of means thereto.

(3) To the industry of the Renaissance scholars' in collecting and copying MSS. we are indebted, in the case of Virgil, for our best materials for the text (see above sect. A, on Codd. Vat., Med., Pal., Rom. and fragm. Aug.), and in the case of many writers for its very existence. Thus Poggio Bracciolini, while attending the Council of Constance in 1414 as Apostolic Secretary, explored the libraries of Swiss and Swabian convents, discovering many lost works, and full texts of others hitherto only known in mutilated copies. At St. Gall he found an entire MS. of Quintilian, of whose writings only half had been known before in a corrupt text. This he copied with his own hand, and also MSS. of Lucretius and Columella and to him was owing the discovery of Silius Italicus, Vitruvius, Manilius, Frontinus, and others. This work of collecting had been begun by Petrarch and Boccaccio; and it was carried on throughout the 15th century, especially under the patronage of the Medici at Florence. The fruit of their enlightened liberality was the famous 'Laurentian ' library, founded by Cosmo de Medici about 1440, enlarged by his son Piero, and further enlarged by (and named after) the great Lorenzo, who swept the monasteries of Greece for Greek MSS. The chief work of learned men at Florence during the age of Lorenzo was the elucidation and correction of ancient texts, written by ignorant and careless copyists, or obscured by dirt and damp in neglected lumber-rooms of monasteries. Until the invention of printing (1450) fixed the text of whole editions, the critical amendments of such scholars could only be made useful through their oral lectures; notes taken at which are probably the basis of many of the diffuse commentaries found in the older editions. The accepted notion that every great poem must have a hidden meaning encouraged the allegorising

1 For accounts of the Italian 'humanists,' and the revival of classical learning, see Symonds, 'Renaissance in Italy' (Smith, Elder & Co. 1877); Hallam, Literature of Europe,' Part i.

tendency which marks Renaissance commentaries on Virgil; and later ages have had to sift the wheat of grammatical and antiquarian lore from much chaff of allegory and mysticism. Orthography (see below sect. III.) was purified of much accumulated barbarism and the efforts of the Renaissance scholars in this direction deserve all praise, though more recent knowledge has detected flaws in the 'conventional' system which they introduced and which has held its own almost to the present day.

:

Among the scholars of the age of Lorenzo the leading names are those of Pico della Mirandola (1463-1494), the gifted exponent (and victim) of the fancies of the Hebrew Cabbala; Marsilio Ficino, the student of Platonic philosophy; Cristoforo Landino (1424-1504), Professor of Latin literature (Rhetoric and Poetry), a copious annotator of Virgil and Horace, though now better known as an editor of Dante; and Angelo Poliziano, whose name is as representative of letters in the last half of the 15th century, as that of Poggio in its earlier part, or Cardinal Bembo in the first half of the 16th century. Poliziano (i. e. of Montepulciano, Latinised to 'Politianus' and re-Italianised), born in 1454, became Professor of Greek and Latin at Florence about 1480. His fame as lecturer, scholar, and writer of Latin poetry eclipsed all his contemporaries, and gave him a virtual dictatorship of the world of letters. His connection with Virgilian learning has been alluded to above (sect. A. p. xxiii). Equally eminent as scholar, critic, and writer of Latin poetry was Pietro Bembo (b. 1470, d. 1547), a noble Venetian, secretary to Pope Leo X, after whose death in 1521 he retired to a literary and learned life at Padua, where he accumulated a rich library: but in 1539 returned to Rome on being made a Cardinal by Pope Paul III. The Cod. Vaticanus of Virgil was once his property (p. xxii); but the celebrated 'Cod. Bembinus' of Terence (4th or 5th century) in the Vatican Library takes its name from one Bernardo Bembo, probably a relative.

Collection of materials for future criticism was thus the chief work of the Renaissance scholars: their own restoration of Latin texts being liable to much uncertainty. Thus, their aim being to restore a readable text rather than the 'ipsissima verba' of the writer, they often fill up lacunae in MSS, especially in the case of poets, where their own practice in Latin versification tempted them to supply what was lacking. The text of Virgil was independent of their aid; but more corrupt texts were frequently interpolated by them: see e. g. Munro on Lucretius (Notes I.) iii. 89, iv. 532 and Introd. p. 9; and cp. Tibullus i. 2. 25 ('En ego

cum tenebris tota vagor anxius urbe '), the missing pentameter of which was supplied in four different ways, (1) by Aurispa (1400) 'Securum in tenebris me facit esse Venus,' (2) by Thomas Seneca (1420) Praesidium noctis sentio adesse Deum,' (3) on Cod. Vat. (end of 14th century) 'Ille Deus certae dat mihi signa viae,' (4) 'Usque meum custos ad latus haeret amor'—all respectable substitutes.

E. Printed Editions 1.

Some idea of the rapid multiplication of Virgilian texts by the agency of printing may be formed from the fact that, from the appearance of the 'Editio Princeps' at Rome in 1469 until the year 1600, there are only seven years in which no fresh edition of Virgil is recorded, while many years are credited with two, three, or more; mainly from the presses of Venice, Milan, Florence, and Rome, though in the 16th century we find mention of Paris, Leipzig, Antwerp, Amsterdam, and London 2. The first English edition whose date can be determined appeared in 1512 from the press of Wynkin de Worde, and was followed by those of 1514, 1516, and 1533: but mention is made of a 'vetus Londinensis editio,' published by Pynson, a printer in London between 1493 and 1528, which may have been older.

The 'Editio Princeps' was published at Rome in 1469 'sub auspiciis Paulli II. in domo Petri de Maximo per Conradum et Arnoldum Teutonicos,' and was repeated in 1471; 550 copies in all (275 of each issue) being printed, of which (according to Heyne and Wagner) only six were known to exist in 1832, the best being in the library of St. Geneviève at Paris. The editions of the 15th century, though in some cases professing to follow good MSS, are but of slight critical value. Several editions appeared from the Aldine press at Venice in 1501 and subsequent years. Among

1 The information here given is mainly derived from vol. iv. of Heyne's Virgil (ed. Wagner, 1832), pp. 635-749 ' De Virgilii Editionibus.'

2 The British Museum Catalogue of printed books (1882) under the heading 'Virgilius Maro' (editions and commentaries, translations, and general literature of the poems) occupies 74 folio columns. It specifies 220 editions: viz. 15th century, 18 (mostly Italian); 16th, 65, including the 'first Aldine' of 1501; 17th, 33; 18th, 48; 19th, 56. Of translations of the whole works it has English 21, French 18, Italian 9; also German, Polish, Spanish, and Dutch, and many translations of portions of the poems: with such curiosities of Virgilian literature as travesties, parodies, centos, etc.

later editions of importance were those of De la Cerda (Madrid 1608–1617), a commentary in 3 vols.; Nicholas Heinsius (Amsterdam 1676); Burmann, with comments of Servius, Philargyrius, Nicholas Heinsius, etc. in 4 vols. (Amsterdam 1746); Heyne, in 4 vols. (Leipzig) 1st ed. 1767-1775: 2nd 1788: 3rd 1798–1800 (in 5 vols): 4th, edited by Wagner, 1830-1832, including Philip Wagner's' Quaestiones Virgilianae' (vol. iv. pp. 383-587), the foundation of much recent Virgilian criticism. The title of this Heyne-Wagner edition is 'P. Virgili Maronis carmina ad pristinam orthographiam revocata'; and its text represents the first systematic attempt to rescue the text of Virgil from the errors of the 'conventional ' system of orthography introduced by the Italian scholars of the Renaissance, the principles on which it proceeds being expounded in Wagner's Orthographia Virgiliana' published in 1841. Of this work Mr. Munro says, 'With admirable industry he (Wagner) amassed all the evidence afforded by the Medicean and, so far as it was accessible to him, of the other ancient MSS. of Virgil. As these, like other old MSS, are as a rule very tenacious of the true spelling in those cases where there is only one right method, he performed this part of his work with eminent success, and still remains one of the best authorities on the subject. In those other cases, however, . . . in which variety is the rule of the ancients, and which include a great multitude of particular instances, he has chosen to abandon the safe ground of evidence and experience and has made Virgil write what he decided on a priori principles he must have written. This seems to me the reason why his system was not more generally followed." Wagner's orthography has not, perhaps, been followed in Germany: but in England the preeminence in recent years of Conington's edition (in which it is reproduced with but little change) has made it the standard orthography for Virgil; Ribbeck's elaborate collation of MSS. being familiar only to a comparatively small circle of scholars in this country. In Part III. of this Introduction some reasons have been given for departing in some respects not only from Wagner's system as represented by Conington's text, but from some of the more scientific results of Ribbeck's accumulation of MS. evidence.

The following enumeration of the points treated of (some of them at great length and with much minute illustration) in Wagner's 'Quaestiones Virgilianae' will show how wide and various was his study of Virgilian usage:

i, ii. On the forms 'ab,'' ex' before consonants.

iii. Accusative of proper names, '-an,' '-en' (Greek), ́ -im,' not ' -in,' etc. iv. Greek forms-os,' '-on' ('Arctos,' 'Spercheios,'' Aegyptos,' etc.) v. Termination of 3 plur. perf. act., ' -ere' preferred to '-erunt,' where the metre was unaffected, e. g. at the end of a line.

vi. Confusion of Moods and Tenses.

vii-x. Interchange in MSS. of present and perfect; of sing. and plur. in verbs; of case and number in nouns; of 'ad' and 'in."

xi, xii. Hiatus; and lengthening of short syllables in arsi; each exhaustively illustrated.

xiii. Caesura or pause after first foot; e. g. G. iii. 317, Ecl. i. 24, Aen. vi. 590, etc.

xiv, xv. Omission or insertion in MSS. of 'in,' ‘a,' ‘e or' est.' xvi. Readings derived from old commentators, especially Servius. xvii-xxii. Use of pronouns 'is,' 'ipse,' 'iste,' 'hic,' 'ille,' 'quis' or 'qui' interrogative.

xxiii-xxviii. Use of 'hic' (adv.); ‘iam'; 'tam,' ' tunc'; 'adeo'; 'primus,' 'primo,' 'primum.”

'ultro ';

xxix-xxxi. Use of participles; absolute (historic), infinitive; interroga

tions.

xxxii. 'Nec,' 'neque,' when respectively used for rhythmical reasons. xxxiii. Apposition and epexegesis.

xxxiv, xxxv. Use of the copulas et,' ' ac' ; ' que,' ' atque.'

xxxvi, xxxvii. Use of particles' ve,' 'at.'

xxxviii. 'Natus,' 'gnatus' (the more solemn form).

xxxix. Anachronisms; e. g. Aen. i. 427, ii. 503, v. 113, vi. 69, 366, 430, X. 449, xii. 269.

xl. 'Virgilius dormitans aliquando,' (1) in 'epitheta ornantia' (e. g. Ecl. v. 27, G. iii. 343, Aen. xi. 510); (2) inconsistencies (e. g. Aen. ii. 567 sqq. and vi. 523 sqq.; x fin. and xi init.); (3) mistakes (e. g. 'Medus Hydaspes' G. iv. 211).

xli. 'Epulae repostae' G. iii. 527.

It is unnecessary to do more than refer briefly to more recent editions. That of Forbiger (1836-1839, 2nd edition 1845, 3rd 1852) follows Wagner's text; the notes being a useful compilation in which previous editors have been laid under contribution, and which has in its turn been laid under contribution by Professor Conington (1st ed. 1858-1871, 4th 1881); whose commentary, completed by Professor Nettleship, is now the standard authority on Virgil in England, and is not likely to be easily superseded. More important than any of these (perhaps than any other previous edition whatever) for the textual criticism of Virgil is that of Ribbeck, containing as it does a more complete 'apparatus criticus' than had hitherto been accessible, with probably the first accurate collation of one most important MS, the Codex Palatinus. Vol. i (Bucolica and Georgica) appeared in 1859; vol. ii (Aen. I–VI) in 1860; vol. iii