of Mr. Maskelyne, Professor of Mineralogy at Oxford, who has permitted me to search for whatever best suited my requirements in his very extensive collection of electrotypes, formed with equal taste and judgment from the choicest gems in every important cabinet, whether national or private, now in existence. Lastly, it is necessary to mention that some of the cuts, taken from gems in my own possession, have made their appearance in other books, but are introduced here when nothing equally appropriate to the place was obtainable from any other quarter; and also that wherever no name is appended I have been unable to ascertain the ownership of the intaglio from which that particular cast was taken.

UNIVERSITY

CALIFORNIA

TH

INTRODUCTION

`HOUGH it would be out of place in an edition like this to dwell at any

on

to dwell at any length on the criticism of our author, yet in justice to myself a few remarks must be made.

It will hardly I think be disputed by competent judges that Horace is one of the most difficult Latin writers, almost or quite as difficult as Virgil by the intrinsic qualities of his genius and his style; while the obscurity is accidentally increased by the sources being far more troubled from which his text has to be derived.

Not a single manuscript has survived the deluge that swept away the old classical world: the most ancient are posterior to the great Carlovingian revival. Italy appears to possess none of any great age or authority: the best so far as is known of those now in existence are found in the libraries of Switzerland and Paris; though the most valuable of these have been but recently collated, and many of them, for one half of Horace at least, are still unknown to us. The most important English manuscript belongs to Queen's College Oxford. Bentley found it of much use in his edition through the great kindness of Mr. Nettleship of Lincoln College I am in possession of a fuller collation of it. It appears to me to be equal to any of the Swiss or Paris manuscripts, with the sole exception of Orelli's oldest Bernese; and it is often tenacious of the genuine spelling, as in mercennarius pilleolus Zmyrna, &c. ; some of which so far as I know appear for the first time in the present edition.

It was in 1837 and 1838 that Orelli published his popular edition, with the readings of the oldest Bernese and other

Swiss manuscripts. These however were not given with the accuracy and completeness which are now required. The chief merit of Ritter's edition published in 1856 and 1857 is a fuller and more accurate collation of these same manuscripts; to which collation I have had to trust entirely in the satires and epistles. In 1864 Mr. O. Keller gave to the world his revision of the odes and epodes, accompanied by a most elaborate critical apparatus, a collation of the Paris manuscripts of the tenth and later centuries, the readings elicited from Horace's scholiasts, and the Testimonia Veterum'. We find also in this edition a collation by Mr. H. Usener of the oldest Bernese, which professes to be more accurate than Ritter's, with what truth I am unable to say. Keller's colleague Mr. A. Holder has not yet given us the satires and epistles. This I much regret for other reasons, and because we are promised a fuller account there of the codices employed in both portions of the work.

Keller's theory of the relations of the several manuscripts and the readings of the scholiasts to one another, and of their comparative importance, is so complicated that I can hardly realise it to myself. Agree with it I cannot; for it leads him to place in the more interpolated classes the oldest Bernese and the antiquissimus Blandinius'. The general opinion of late critics seems to me indisputably true, that this Bernese is far the best of existing codices: it appears to be at least a century older than any of the rest; and it was just in that interval that designed interpolation began to be rife. Unfortunately it is very incomplete: it wants twelve whole odes and portions of more than twenty other odes or epodes: of the rest of Horace it contains only the first three satires and the Ars.

I do not pretend to decide on the relative merits of the other Swiss and of the Paris MSS.; nor would it I think be easy to do so. But no editor of Horace can forbear to offer an opinion on the long-lost Blandinii' of Cruquius, by which the criticism of Horace is at once so greatly complicated and

so immensely benefited. Since Bentley all or almost all the most sagacious scholars have looked upon what Cruquius names his antiquissimus Blandinius' as by far the most precious recorded manuscript of the author; and with them I entirely concur. No one I believe now doubts the essential good faith of Cruquius; but in his days the art of collation was ill understood, and we are often left in total ignorance, often in doubt as to what his manuscripts really contained. The antiquissimus' is much more unerring in the satires and epistles than in the odes: in the latter its readings seem often to be interpolated, though on the whole superior to those of any other codex. Yet so far as I am able to judge there are many passages, generally involving however but a single word or so, where it would be impossible to say whether it has the true or the false reading, unless we had some higher authority able to control and overrule both it and all our other existing authorities.

For English readers I need scarcely say that, with one memorable exception, I do not believe in the spuriousness of any existing poem of Horace or of any portion of any one of them. There is not a tittle of outward evidence for such a supposition, either in any manuscript or in any scholiast or grammarian. Horace for a man of his powers I look upon as a very unequal writer, and many of his poems I do not rate very highly. But his style throughout is his own, borrowed from none who preceded him, successfully imitated by none who came after him. The Virgilian heroic was appropriated by subsequent generations of poets, and adapted to their purposes with signal success: the hendecasyllable and scazon of Catullus became part and parcel of the poetical heritage of Rome; and Martial employs them only less happily than their matchless creator. But the moulds in which Horace cast his lyrical and his satirical thoughts, were broken at his death. The style neither of Persius nor of Juvenal has the faintest resemblance to that of their common master: Statius whose hendecasyl

lables are passable enough, has given us one alcaic and one sapphic ode, which recal the bald and constrained efforts of a modern schoolboy. I am sure he could not have written any two consecutive stanzas of Horace; and, if he could not, who could? And then to suppose that forgeries of whole poems or portions of poems should have taken place, and yet left no trace of the fact behind them, during ages when he was in every public and private library! For that they certainly must have had their origin at an early period, say before the time of Quintilian, is admitted I believe by all judicious advocates of such forgeries: to give them a later date would lead to endless contradictions and absurdities.

That the custom of fathering spurious works on famous writers was prevalent in those days is notorious enough. All know the minor poems attributed to Virgil. Now whatever others may think, I hold it to be a demonstrable truth, though this is not the place to enlarge upon it, that Virgil, to take one sample of them, never wrote any Culex whatever; that the Culex we now have was forged little if at all later than the time when the Aeneid was published, by some one who wished to adorn with a Batrachomyomachia the apotheosis of the Italian Homer; and that this is the very poem which was accepted as Virgil's by Lucan when he exclaimed 'et quantum mihi restat ad culicem'; by Statius when in the preface to his Silvae he writes sed et culicem legimus et batrachomyomachiam agnoscimus, nec quisquam est illustrium poetarum qui non aliquid operibus suis stilo remissiore proluserit' (a good commentary on the causes of its origin); by Martial, Suetonius, and others. And so Suetonius speaks of having had in his hands elegies and letters attributed to Horace, but in his opinion spurious. Surely however forgeries like these, common in modern as in ancient times, lend no support to the theory that long spurious passages could have been introduced among the genuine works of Horace and Virgil during the only ages in which those forgeries could possibly have been composed.