supposed orthography of the poets of the Republic, but their rule was purely arbitrary, and founded on conjecture instead of analogy. We are only now beginning to acquire a body of rules, drawn from those manuscripts which are thought to represent the fashion of the best ages.

It would be easy to multiply instances of ignorance on the part of monastic copyists. The text of Pliny has suffered particularly from two causcs; one, his frequent use of proper names and technical terms, which led first to errors of misconception and then to interpolation; the other from certain peculiarities of idiom and construction, such as the favourite employment of the ablative absolute. Mistakes like the substitution of inscitia for in Scythiâ may possibly have arisen from dictation; and a study of these might give some clue to the early pronunciation of Latin. The elaborate beauties of Ciceronian Latin, the ȧxpíßeia, as Ernesti calls it, and the rhythmical correspondences of his diction, when once the ear is familiarised with them, have enabled men like the last-named critic to detect frequent errors of transcription, from the ignorance of copyists, who have transposed single words, or omitted others which they thought redundant, thus marring in reality the just measure of the sentence. The changes and corruptions of the Latin language itself, even in Italy, added to this source of error. These corruptions had already begun in the first centuries after the Christian era; and although a better prose style began with Claudian, which lasted from A.D. 350 to 500, the classical age of Christian literature, yet Ducange has clearly proved, with others, that writers of that epoch, whose faults would be reflected in the colloquial Latin of the copyists, corrupted the Augustan purity of the language by adapting the phraseology to new uses, based upon different conditions of life and habits. Barbarisms crept in as early as the third century, and the Latin language had lost its purity, and was fast losing its genuine idiom in the fourth. Attic Greek was only faintly imitated at Constantinople at the same epoch, and the manuscripts written there in later days bear abundant traces of barbarism. The best copies of Cicero were made in Germany, but Ernesti has traced a multitude of corruptions to the foreign and alien vernacular.

Caligraphy bred a host of depravations from the foolish desire of subordinating the accuracy of the text to superficial embellishment. Give me,' said Jerome, as early as the fourth century, non tam pulchros quam emendatos codices,' and John Gerson describes a class of monastic scribes as 'quasi pictores, who had absolutely no knowledge of what they were tran

scribing. Excessive care for mere external elegance and the form of single letters distracted their attention from the context, even where the language was understood. The profits, both of copyists and correctors, when the office became one of gain, was enhanced by the appearance of neatness, to which fidelity was accordingly made subservient; and hence critics concur in remarking that where the writing is more beautiful, more errors commonly remained. Nor was the remedy, if the copyist discovered the error or chose to correct it, much better than the mistake itself. Interlineal insertions, as a rule, were carefully avoided: if the accidental omission, even of a single syllable, took place, the whole sentence was repeated, or the omission perhaps inserted in quite a foreign place, in order not to spoil the look of the manuscript; minute symbols being sometimes added to avoid erasure, which the next copyist would not notice or would purposely omit, or which became obliterated by time, leaving the original error apparently uncorrected.*

But the worst mistakes are those where the copyist has endeavoured to extract a meaning from what he could not understand; and, unfortunately, the sciolism of scribes and correctors seems to have kept pace with their ignorance. Instances abound where copyists would appear to have exhausted their ingenuities in depraving the text and obliterating the previous vestiges of the genuine reading contained in more. clumsy corruptions. The twenty-second book of Livy began thus: Jam vero appetebat, quæ Annibal ex hybernis metuit, et neque eo qui jam ante conatus transcendere Apenninum, until Valla discovered in the words italicised the true reading, nequaquam. The other occurs in Cicero's Philippics,' xiv. 3, where the conjectural, but obvious, improvement of non 'injustæ belli internecini nota' has superseded the nonsense contained in MSS. non injusta evelli inter nec uno te.' We might add a similar example in Pliny's Natural History,' vii. cap. 7; where the corruption of tamen cujus semper tinctoria (altered marginally into in victoriâ) est mens,' was restored with much probability by Rhenanus into curâ pertinaci æstuans.'

Mr. Taylor alludes to a distinct source of corruption in the intentional omissions or alterations of fraud, which, so far as they can be proved to exist, undoubtedly affect the credit and value of the text.' Griesbach, whose canons of critical emendation form the basis of all interpretation of Scriptural

* See Gesner ad Horat. A.P. 45, where Bentley's sagacity detected a transposition of this kind.

VOL. CXXXVII. NO. CCLXXIX.

G

MSS., and deserve the careful study of the classical student, mentions, among other readings of suspicious authority, those which inculcate precepts of monastic devotion, and appear to support so-called orthodox doctrines; and the text appears to have suffered considerably in much earlier times at the hands of religious sects and factions, especially of the Eastern Church.* But the traditionary instincts of Romanism, with which monastic scribes were imbued, did not extend to any serious corruption of classical literature, preserved through their means. In Lucan's Pharsalia,' vii. 725, the expression, 'Plurimaque in savos populi convicia divos,'

appears to have offended the pious cars of some copyists, who have omitted the line from their transcript; and v. 796 of the same book,

'Fortunam superosque suos in sanguine cernis,' absent also in some MSS., may, very possibly, as Weber conjectures, have been purposely erased by some superstitious monk, who misunderstood and therefore misinterpreted the meaning of the words in italics. We know the rule of St. Isidore hereticorum libros legere nefas,' and we are told that Jerome dreamed he was whipped by the devil for reading Cicero; but the condition of classical MSS. forbids the idea that the hostility of the early Fathers, or even of Gregory the Great, preferred the subtle method of poisoning the purity of the text to the policy of open prohibition. The lacunæ common to all existing copies-especially in Book III. of Cicero's 'De Naturâ Deorum-are ascribed to the mutilations not of Christians but of heathens, who saw in the candid exposure of the follies of Pantheism abundant arguments in favour of Christianity. Guyetus denied the authorship of the last four books of the Eneid on very insufficient grounds; but not all the wildness of allegorical interpretation-which misrepresented rather than corrupted the text of Virgil-will give colour to the extravagant scepticism of Hardouin, who ascribed the entire poem, in common with the mass of Greek and Latin authors, to some spurious productions of the thirteenth century.‡

* Compare Erasmus, pref. to ed. Nov. Test., 1535.

† Fabricius (Not. Lit. de scriptis Ciceronis philosophicis) cites Arnobius' treatise Adversus Gentes,' iii. 103. Erant,' says that author, qui mussitarent, oportere statui per Senatum Romanum, 'aboleantur ut hæc scripta, quibus Christiana religio comprobetur, et 'vetustatis opprimatur auctoritas.'

He admitted the genuineness of part of Horace and Virgil, but

:

The varieties and corruptions of even the oldest MSS. serve to explain the early origin of critical studies. It was the singular fortune of the Homeric poems to develope a kind of philological criticism before even the name of critic or grammarian was attached to a profession. Two emendations of a sophist under Pericles, observed by Wolff,* show how vicious were the copies at that early period; and H. Stephens infers the same of the text of Eschylus from the frequent resort of the old scholiasts to conjecture. We must briefly turn from the necessity of criticism to its methods, to show how the purity of classical authors has suffered from incompetent emendation. Philemon, in his Homeric Questions,' attributes much of the errors then existing in the copies of Herodotus, Thucydides, and other Greek writers of eminence to the Tapadiopθώματα πάνυ ἄγροικα—the rustic corruptions of mistaken critics and even in the days of pure Latinity we learn from Gellius that the Roman classics were deformed, not only by the casual mistakes of copyists, but by the deliberate perversions of falsi et audaces emendatores.' We may readily concede that in many instances the earliest critics were in possession of older, and to that extent better, MSS. than any now surviving. On the other hand, chance has preserved to us some monuments of antiquity, as Villoison's Codices Veneti of Homer, evidently unknown to Eustathius, which in the days of private study and restricted intercourse were inaccessible to the compilers of new recensions.' Several Latin authors appear to have found interpreters soon after their works were published. Glosses, or short explanations of difficult or unusual words, were then mostly in vogue; they were written at first interlinearly, and afterwards in the margin, until they extended in the twelfth century to a kind of running commentary. The evils of this practice consisted in the temptations offered to later copyists or more indolent critics, who neglected the reading of the text for the more familiar annotation of the scholiast. If not actually

described them as two allegorical writers, who, as Lalage and Æneas, represented Christianity and the life of its founder. The Odes he ascribes to some pseudo-Horatius in the latter half of the thirteenth century. His explanation of Od. ii. 20, Non usitata, etc.,' is given by Gesner. Prosopopeia hæc est Christi-triumphantis et Judæos alloquentis. Biformis vocatur Christus, quia simul in formâ Dei et in forma servi. Alitem album interpretatur "candidâ veste indutum:" -Que residunt pelles cruribus aspera-ocreas intelligit, quibus equitans Dominicanus crura tegit!!!'

Proleg. in Hom., cap. xxxvii.

† Noct. Att., ii. 14.

6

A

embodied into the text, and as in some instances of Roman poets, put into metre accordingly, these exegetical or explanatory phrases acquired too often a spurious authority as various readings. The relics of Acron and Porphyrio, the ancient Horatian scholiasts, abound in interpolations; while the authoritative readings of Servius, based in all probability on excellent materials, are buried in a mere farrago of commentary, the incrustations of later and less scrupulous physicians of the text. The concise diction of Tacitus was peculiarly obnoxious to this source of corruption, from the necessity of explanatory glosses to readers enervated by the common-place simplicity of Suetonius or Eutropius. The value of the early glosses, which we are far from wishing to impugn, depended on their preservation by later copyists or critics as distinct from the text itself. commentary on Lucan still exists, entire in twelve manuscripts, and compiled by Vacca, a grammarian of the sixth century, from a multitude of much older glosses containing numerous readings of genuine mark and high antiquity; and the Corpus Glossatum' of the Roman law, the labour of Accursius in the thirteenth century, has elicited the warm eulogium of Savigny on the critical industry of the early jurists in the formation, as well as the interpretation, of the text. If no other good were done by such later compilations as the one just mentioned, they deserve acknowledgment for having restrained the luxuriousness of interpretation which had so long prevailed in the schools. The laborious triflings of the centonists had early corrupted much of the purity of Virgil, especially in his minor poems, as well as of Ovid and other favourite poets. A monkish interpreter in the Codex Gudianus of Virgil derives Publius 'a pollice magno, quem habebat,' and Virgilius à virgâ lauri,' but it would be well if the pedantic subtleties of the rhetoricians had been confined to mere fantastic derivations. The arbitrary interpolations of scholastic teachers corrupted the text in a wholesale fashion under the pretence of emendation; and manuscripts, like the Codex Bernensis of Cicero's Tusculan Questions, or the copy on which the editio princeps of Terence was founded, and which was revised, according to the colophon, by Calliopus, a Ma'gister Scholasticus' of the time of Charlemagne, exhibit the same traces of depravation. The reason why Plautus is less injured than Terence is because the former has partially escaped the perverse ingenuity of these critics; and we find Merula, the first editor of Plautus, thus explaining the inequalities of his MSS. :- At septem ultimæ Comœdiæ,' he says, ut in eas incidimus, quæ simplices et intacta a censoribus

6