and where the first editors break silence, it is but to join in the Jeremiad over the paucity and corrupt character of their materials.* Much allowance must be made for what Froben stigmatises as the servile adherence of first editors to one particular manuscript. The niggardness of bibliophagists is a frequent theme of complaint, though perhaps exaggerated; but while some private patrons of literature, as Navagerius of Venice, to whom Aldus dedicates his Pindar, appear to have lent a MS. in their possession to this or that scholar, who was to superintend its printing, yet, even if the copy was not unique, fidelity of editorship was in some cases the condition attached to the loan, and precluded the editor from fishing in other waters. The utmost that some of them appear to have attempted was to produce a faithful facsimile in print of the MS. they employed. The first edition of Terence contained no divisions into verse: that of Asconius Pedianus reproduces the blank spaces existing in Poggio's copy; the undated editio princeps' of Horace appears to have been the mere mechanical production of the typographer, unspoiled by the ingenuity of corrector or grammarian. Gesner, accordingly, like Reitzius with Lucian and Burmann with Claudian, describes the text as equal to the authority of a MS. The state of these MSS., like those of other authors, can best

เ

* Cf. J. Andreas' preface ad Virgil. In tantâ tamque mendosa exemplariorum raritate;' and again:-'Ecce emisimus Ciceronis ad 'Atticum opus, et multis in partibus, quod notis secretioribus, et ' inter paucissimos cognitis, scriptæ sunt literæ, non satis intellectum ; in parcitate præsertim exemplariorum, quæ aut non sunt apud literales 'erogandi, aut at invidis communi hominum odio occultantur.' Aldus (ad Aristot. 1495-8) could find only one copy in Italy of Theophrastus or of Aristotle's 'Moralia;' and a private patron lent him a MS. for Hesychius, the only one on which that author depends. Of Dion Cassius R. Stephens says: In iis libris (i.e. twenty-three out of the eighty mentioned by Suidas) excudendis, unico exemplari usi sumus, eoque accuratam et diligentem castigationem desiderante.' Ferandus of Brescia speaks in identical terms of the single copy on which his 'editio princeps' of Lucretius was based; a preface strangely omitted by Mr. Botfield.

† Gerard Falkenberg in his preface to Nonnus, addressed to Sambucus, contrasts the liberality of his patron with the jealousy of others: -In Italiâ præsertim et Gallià, qui, si quæ habent veterum codicum exemplaria, vel sibi ca, ut soli sapere videantur, reservant, vel non'nisi carissime vendita typis describi patiuntur.' John Andreas (ad Lucan.), like Ugoletus (ad Quintil.), complains of the stupidity, avarice, or jealousy of private owners in withholding their MSS., because they thought that printing depreciated their value.

6

ex

be judged by the editions themselves, and the text will supply the information which Mr. Botfield's Prefaces fail to convey. The study of these editiones principes' has been too often undeservedly neglected even by modern critics; the prestige of the Aldine press threw many of them into the shade, although both that and the Juntine did not scruple to pilfer without acknowledgment from their contents. Nor can much importance be attached to the statements of later editors of the sixteenth century. Henry Stephens, for example, in his edition of Herodotus in 1570, speaks of his labours as vetustis exemplaribus recogniti,' but is forced in his second impression to confess that, up to that time, he had not been able to consult a single ancient MS., and was probably content to utilise the collation of Aldus. Dolus latet in generalibus; and their loose and ambiguous professions of the number and antiquity of their MSS. deserve small attention. Some undoubtedly had access to excellent and ancient copies, since lost. Valla speaks of a Codex Regius of Livy, 'quo nullus in omni Italiâ est augustior,' which only survives in his readings. Lipsius, who is singularly reticent as to his authorities for Tacitus, had full opportunities for ransacking the Vatican; and Victorius, himself a diligent collator, was the first Ciceronian editor who could boast of using any MS. of his Epistles ad Fam. earlier than the fourteenth century. On the other hand we meet frequently with feeble and pedantic attempts to give a forced additional value to a MS. of undoubted authenticity. Aldus speaks of his copy of Pliny as coeval with the author; but the partial refutation by Hearne is sufficient to expose the utter absence not only of proof, but of plausibility; nor need we seriously examine the assurance of Zarotus, in his Milan edition of Virgil, that it was corrected ab ipsis propriis Maronis exemplaribus.'

The fragmentary, no less than the corrupt character of their materials will frequently account for the inferiority of the first printed editions. Even where the archetype had exhibited the author in his integrity, his works were too often disjointed by literary men in later transcripts. No entire MS. has been known to exist of the extant books of Livy: the editio princeps,' of Cicero's De Officiis by Fust, contained a leaf with Horace, od. iv. 8, attached, the earliest impression of any portion of that author. The popularity of classical writers had in fact not always extended to the whole of their works, even before the revival of learning. Cicero was probably never studied collectively in the schools of the early grammarians, but through the medium of rhetorical compendiums

serves to

and select passages. An extract from Fronto * explain the patchwork character of the MSS. of his Epistles, and the reason why so many of those epistles have perished.† Throughout the Middle Ages the absence of general views or ideas of symmetry, which was manifested by the preference given to the abridgments of compendious but inferior writers, sufficed to destroy the integrity of classical MSS. in general, while the text of such school-books as Eutropius was overwhelmed, like our topographer Stowe, by the gratuitous additions of his continuators. The Greek anthology was confined to the disjecta membra, the arbitrary expurgations and interpolations of the monk Planudes in the fourteenth century until Salmasius discovered the older Palatine MS. in the seventeenth. The variegated composition of this class of manuscripts, as with Pliny and Homer perhaps in particular, has seriously increased the difficulties of classification with a just reference to the authority of various readings, inasmuch as many of them exhibit at once the most striking features of resemblance and of contrast, while the same copy contains within itself discrepancies which can only be explained on the supposition that they were supplied from a totally different source. Hence the unequal and heterogeneous character of many of the first printed texts, especially in the Opera Omnia' editions. Fragments of MSS. were snatched up promiscuously wherever they could be found, or as chance favoured the editor, to form a collective whole; and good, bad, and indifferent were boiled down together without method or discrimination. John Andreas speaks thus in his preface to Apuleius in 1469: Lucium Apulcium, ut in exemplariorum penuriâ licuit, redegi in unum corpus, variis in locis membratim perquisitum.' Minutianus' Opera Omnia' of Cicero, like Beroaldus'

6

6

* Memini me excerpisse ex Ciceronis epistolis ca dumtaxat, quibus inesset aliqua de eloquentiâ vel philosophiâ vel de republicâ dispu'tatio' (ed. Rom. p. 150). Nerlius says of Homer in his preface to the first edition: Ob incuriam librariorum ita sui dissimilis videbatur, ut in nullo fere codice, quamvis perveteri, integer agnosceretur.'

We learn from the old grammarians that they were compiled originally in separate collections under the names of the persons to whom they were addressed; and very likely they retained this arrangement until scattered by barbarism. Dr. Tunstall assigns their present shape to some collector of the twefth or thirteenth century. Suetonius was not divided into books till the time of Isaac Casaubon. Sigonius attributes the distribution of Livy into decades to some grammarian, Petrarch to the fastidiosa legentium ignavia;' at all events it was probably not anterior to the sixth century at the latest.

Tacitus, was collected second-hand not from MSS. but from earlier printed editions of portions of his author, which in their turn had been similarly compiled. The editio princeps' of Cicero's Orations appeared in 1471, but must have been merely an embodiment of scattered materials. This much is certain, that if the editor employed an entire MS., it must have been of recent date, since it contained those same orations, except thepro Roscio Comodo,' which Poggio had brought in that century from Germany. And the same defects of age and quality attach to those tria vetustissima '-observe the loose phraseology of even eminent critics of that age- atque emendatissima omnium M. Tullii operum,' of which Lambinus boasted himself the possessor in his preface to Lucretius.

Lastly, the insufficiency of the MSS. used by the first editors of the fifteenth century, amply testified, as we have seen, by their own statements and by the study of their works, is shown even more conclusively by the progress of later research. It is a mistake to imagine that printing established the text in the sense of finality, so long as the discovery of earlier and better MSS. revealed its imperfections; and it is a fortunate circumstance that these discoveries were often delayed until critics and editors were sufficiently advanced to appreciate them. Cratander published, in 1528, Cicero's Epistles ad Brutum' from a MS. at Heidelberg. The extant writings of Livy are due to several discoveries made at different times; -the Codex Moguntinus of part of the fourth decade, found in 1518 in the cathedral church of St. Martin at Mayence, which was assisted by the Bamberg MS. found by John Horrion in 1615;-the Codex Laureshaimensis of the sixth century, found in 1531 by Grynæus in Switzerland after its removal from Lorsch near Worms, and containing books 41 to 45;-and finally, Bruns' Vatican palimpsest of part of B. 91. The Aldine editio princeps' of Eschylus contained only fragments of the Agamemnon': a lacuna was left between vv. 301-1034, and again after v. 1129, which was followed by a mutilated part of the prologue to the Choe

The history and authorship of this MS. is curiously shown by the colophon, partly obliterated, Sûtberti épi de dorostat.' Suitbert was first a monk in Ireland, then an English abbot, and elected an apostle of the Frisians in 693. He preached for two years at Dürestadt, until his removal by Pepin to Kaiserswerder. It is conjectured that he brought this MS. thither from Ireland, and that it afterwards found its way to Lorsch, and thence finally to Switzerland. (Syllabus Codd. MSS. Livii, ex recens. Drakenborch. ed. Twiss, 1841, vol. iv. p. 436.)

phoro. Much as Turnebus improved other portions of Eschylus by the aid of his Codex Ranconeti from the library at Paris, this confusion was not remedied until VictoriustheSospitator Ciceronis,' in like manner, from his having first had recourse to the archetypal codex of Cicero's lettersdiscovered an entire MS. which explained the origin of the defect. The restoration of Eutropius was postponed for forty years after Ignatius' first attempt in 1516, until Schonhovius, a canon of Bruges, made use in his Basle edition of the Codex Gandavensis. Commelin's edition of the Palatine MS., commonly referred to the fourth or fifth century, marked a new epoch in the text of Virgil; that of Diodorus Siculus, in 1539, showed cnly the last five books; but Henry Stephens added ten more shortly after from the Claromontane MS. lent to him by his friend and patron Fugger. Herodotus really owes his first respectable appearance to Gronovius, for the materials used by Valla appear to have been miserably disjointed and incomplete. The Ravenna MS. of Aristophanes was not discovered till 1794, by Dindorf, who attributes it to the eighth century; and three years later the learned Abbé Morelli unearthed from the library at Venice the parent copy of bb. 55-60 of Dion Cassius. We have alluded to the valuable discoveries of Niebuhr and Mai. The Codex Bambergensis of Pliny, written in the tenth or eleventh century, and published by Ludovic de Jan in 1831, is described by Sillig as having completely revolutionised the critical treatment of that author; and ten years afterwards the archetype of all existing copies of Tacitus' Germania' superseded the authority of all previous editions in the shape of the Codex Perizonianus. We shall not weary the reader with further instances; the above sufficiently show how tardy and fitful has been the recovery of classical authors.

The title of Mr. Forsyth's work defines the limited object of his able and interesting lecture, which does not embrace the subject of textual criticism. He has considered the history, not the relative authority of classical manuscripts. So far, however, as we have travelled with him over the same ground, we are the more pleased to find a general concurrence of results, since they have been arrived at independently, the preceding pages having already been in type before his book reached our hands. But we object to his definition of the words authentic and genuine. By authenticity,' he says, 'is meant that the original work was really written by the author whose name it bears; and by genuineness that the account it 'purports to give is bonâ fide and not a forgery.' (p. 7.) The distinction, on the contrary, is precisely the reverse, as Trench,

6