Classical authors'.' To have included mediaeval and scholastic writers, especially St. Thomas Aquinas, would probably have about doubled the task, and indeed it would have been an undertaking for which I should feel even more incompetent than I do for the less ambitious one which I have now ventured to attempt. Moreover this work has been in many respects well performed by others, as, for instance, notably by Hettinger for St. Thomas Aquinas; by Lubin for Hugh of St. Victor; by Marriotto de Gagliole for St. Francis, and also for Aquinas. (2) I have also excluded merely apt illustrations, or parallel passages resulting from coincidences of thought. I have confined myself in the Index here printed to passages which (in my judgement at least, though others may not always agree with it) were likely to have been so far in Dante's mind as consciously, or even unconsciously perhaps sometimes, to modify the form of his language. Such at any rate are the scope and the limits of the enquiry which I have proposed to myself.

One advantage may perhaps be anticipated from the materials or statistics now for the first time collected and tabulated. They will, I hope, enable students to form a more complete idea than was before possible of the encyclopaedic character of Dante's learning and studies, and of the full extent and variety of the literary equipment which enabled him to compose works covering a wider range of subjects than perhaps any other writer, certainly any other very great writer, ever attempted. Our admiration is indefinitely increased when we remember the difficulties under which this surprising amount of learning was amassed; when we reflect that it was in the days before the invention of printing, when books existed only in manuscript, and were consequently very rare and precious, and difficult of access; when there were no helps for study in the way of notes and dictionaries, no conveniences for reference, such as divisions of chapters, sections, paragraphs;

I have indeed made one or two special and occasional exceptions, as perhaps in including St. Augustine and Orosius under this title. Also some of the references to Albertus Magnus and the Arabian astronomers are too important to be passed over without some notice.

above all, no indexes or concordances to help the fallible memory (though, happily, no doubt less fallible then in proportion to the reliance placed upon it); when, finally, we add to all this the consideration of the circumstances of Dante's own life, a turbulent, wandering, unsettled life, one of which we may truly say 'without were fightings, within were fears; one intensely preoccupied with fierce political struggles and anxieties, when 'politics' (if we may use so misleading a term) were a question of life and death to those who engaged in them, and defeat meant, as in Dante's own case, exile, confiscation, ruin. The varied and extensive reading of which Dante's works give evidence would be admirable if it had been exhibited under the most favourable conditions of what we call learned leisure,' and with the help of modern appliances, but under the circumstances in which Dante accomplished it it is nothing less than amazing. Nor are these considerations materially affected even when all allowance has been made for the occurrence of secondhand references and the occasional use of handbooks of extracts and quotations, or 'Florilegia,' on both of which matters we shall have a few words to say presently.

As Mr. Eliot Norton has truly said, 'Dante was born a student, as he was born a poet, and had he never written a single poem, he would still have been famous as the most profound scholar of his times. Far as he surpassed his contemporaries in poetry, he was no less their superior in the depth and extent of his knowledge.' Dante is a striking example of what Mr. A. J. Butler has well termed 'the incredible diligence' of the Middle Ages. We marvel at this in our life of feverish haste, as we do at the infinite patience and leisure of Indian and Chinese craftsmen. The learning of Petrarch is also very remarkable, but the circumstances of his life were much more favourable for its acquisition than those in which Dante lived.

This subject has already, in a partial way, attracted the attention of several students of Dante. In a partial way, I mean, because, although some writers have dealt with the quotations to be found in single works of Dante, and others

have written monographs on Dante's use of particular authors, yet no complete and systematic collection or discussion of such passages has yet appeared. None, I mean (1) covering all the works of Dante; (2) including all the earlier authors thus used by him; (3) embracing not only direct citations, but also allusions and references, many of which allusions are equally certain and even obvious, though not introduced by any formal acknowledgement. Such a collection, so far as concerns Scripture and Classical authors, though not including the wide field of Scholastic theology and philosophy, I have now endeavoured to present in the Index which follows. These statistics being, however 'incomplete,' yet as far as they go 'systematic,' enable us to form a judgement as to the comparative amount of use made by Dante of particular writers a point on which some erroneous statements have before now been made-and also as to the extent or limits of his acquaintance with the writings of an individual author when these are many or various in character: the extent in some cases, and the limits in others, being alike remarkable.

It may be convenient to summarize briefly here the general result. If we include (a) direct citations, (b) obvious references or imitations, (c) allusions and reminiscences, it will be seen that more than 1,500 passages may be found that fall under one or other of these heads. It is obviously impossible to fix precise limits to the class 'c,' partly from differences of opinion as to the certainty of an allusion,' and still more from the fallibility of the memory and the imperfect scope of the reading of any one student, even with all the help to be gained by modern appliances, and after all the labours of others in parts of the same field. It is eminently a case in which * παντός ἐστι προσθεῖναι τὸ ἐλλεῖπον. However, starting from the above total as one likely to be approximately correct, or at least proportionately fair in relation to different authors, we may analyse the result further thus:-The Vulgate is quoted or referred to more than 500 times, Aristotle more than 300, Virgil about 200, Ovid about 100, Cicero and Lucan about 50 each, Statius and Boethius between 30 and 40 each,

Horace, Livy and Orosius between 10 and 20 each; with a few scattered references, probably not exceeding 10 in the case of any one author, to Plato, Homer, Juvenal, Seneca, Ptolemy, Aesop and St. Augustine, if we may be allowed to extend the term 'Classical authors' so as to embrace all those mentioned. Further, we suspect on two or three occasions a possible knowledge of Valerius Maximus, though he is nowhere mentioned by Dante. It is to be again remembered that Peter Lombard, Bonaventura, Hugh and Richard of St. Victor, and, above all, St. Thomas Aquinas and Albertus Magnus, to say nothing of Alfraganus, and possibly other Arabian astronomers, fall outside the limits above proposed, though if they were included the above total would be very largely increased.

Probably what will at once strike most readers with surprise in the above summary is the very small use made by Dante of Horace. And the surprise will be increased when we observe that the quite certain quotations of Horace are only about seven in number, and that of these no less than six are from the Ars Poetica, the only one outside its limits being the passing expression, 'bovem ephippiatum' (which recalls 'Optat ephippia bos' of Epist. I. xiv. 43), occurring in Vulg. Eloq. ii. 1. This is certainly not the general impression, as appears from the following statements of two recent, well-known, and generally well-informed writers on Dante. Dante's prose works supply many quotations from Horace (Convito, passim).' And again, From the frequent quotations in the Convito, it is evident that Dante had a special predilection for . . . the Ars Poetica of Horace.' The conclusion here is more correct than the premises, for there is only one definite quotation from Horace (Ars Poetica, it is true) in the whole of the Convito 1.

6

It is interesting to compare with the results thus tabulated. some passages in which Dante definitely expresses his admiration or preference for particular authors. The best-known of

There is also a passage (iv. 12) where Horace is referred to in general terms, together with 'Solomon and his father,' Seneca and Juvenal, as having proclaimed the deceitfulness of riches.'

these is undoubtedly that which contains the celebrated selection of the five great poets of antiquity, viz. Homer (the 'poeta sovrano '), Virgil, Horace, Ovid, and Lucan; after which Dante, with a splendid audacity worthy of Aristotle's μeyadóvxos, ranks himself as the sixth 'tra cotanto senno.' It is worth noticing that in the Vita Nuova 2, Dante's earliest work, he quotes illustrations of prosopopoeia, to justify his own practice, from just these five poets, in the order, Virgil, Lucan, Horace, Homer, and Ovid; the quotation of Homer being taken from his citation by Horace in Ars Poetica, l. 141 3.

1 Inferno, iv. 88 seq.

2 § 25.

3 It is also interesting to compare with Dante's selection of six poets that of Lord Macaulay, who had of course a wider area of choice. He held the six greatest poets of the world (in the order of merit) to be (1) Shakespeare, (2) Homer, (3) Dante, (4) Aeschylus, (5) Milton, (6) Sophocles. When a plea was put in for Virgil, Macaulay not only refused to recognize it, but expressed the singular opinion that both Lucretius and Ariosto should come before him. With this again we might compare the advice given to a young friend by Erasmus, to avoid inferior literature and to stick to Virgil, Lucan, Cicero, Lactantius, Jerome, Sallust, and Livy' (Froude's Erasmus, p. 26). Also G. Villani, viii. 36, says that he was fired to undertake his History when he was at Rome for the jubilee in 1300 by the example of the works of Virgil, Sallust, Lucan, Livy, Valerius, and Orosius.

Chaucer is probably thinking of Dante's list when in Troilus and Criseyde (Il. 1791, 2) he writes:

6 And kis the steppes, wher-as thou seest pace
Virgile, Ovyde, Omer, Lucan and Stace.'

Here Statius takes the place of Horace as in the passage above quoted from Vulg. Eloq. It is a curious coincidence that in a passage which I lately came across in Rabanus Maurus (d. 856) the same five poets are selected for preeminence, but with a ludicrously different object. In a passage which is probably the most grotesque piece of fulsome flattery to be found in all literature, the author compliments an anonymous poetaster (whose works, it is needless to say, are not enrolled in the book of fame) by proclaiming his superiority to all those 'che anticamente poetaro,' and in particular just these five poets are singled out for unfavourable comparison with this new light. The whole passage is so curious that it is worth transcribing :

'Carmina nempe tua dico meliora Maronis
Carminibus, celsi cantibus Ovidii,

Odis quae cecinit Flaccus, verbosus Homerus (!),

Corduba, quem genuit, Africa quem tenuit. (Lucan?)

Hi quia protulerunt pomposis falsa Camenis

Rite tabescentes morsibus invidiae:

Tu devota piis connectis vincula verbis,' &c., &c.

(Ed. Migae, iii. p. 1588.)