in a réchauffé of Aristotle, De Gen. Anim., until Dante has to part company with Aristotle's guidance, and to follow Aquinas and other Christian teachers in expounding a Creationist theory of the origin of the soul (l. 67 seq.). When Dante raises the difficulty which leads to this 'revela'tion 'as its solution, Virgil deliberately stands aside and gives place to Statius, who apologises for assuming the office of a teacher while Virgil is present (ll. 28-33). We have called it a 'revelation,' for such the exordium of Statius implies it to be.

'Se la veduta eterna gli dislego' (1. 31; comp. 1. 36).

We cannot offer any explanation of the selection of Statius for this purpose; nor do we see what there was in his character, history, or literary qualities to render it in any way appropriate. He was, indeed, in one degree (as Scartazzini points out) more appropriate than Virgil, because, being supposed a Christian poet, Christian doctrines could with less violence be put into his mouth. But this qualification is itself a pure fiction of Dante's own, and does not, therefore, much advance the solution of the main question.

4. Finally, What, it may be asked, does Statius symbolise? The part assigned to him is almost, if not quite, as conspicuous as that of the still more mysterious personage Matelda in the later cantos of the Purgatorio, and only surpassed in importance by the parts played in the action of the poem by Beatrice and Virgil. This naturally disposes us to expect some definite piece of symbolism. It is easier to feel convinced here of the reality of a problem to be solved than of its solution. We venture tentatively and with much hesitation to offer the following suggestion as perhaps worth considering faute de mieux. If Virgil (as is generally admitted) represents Human Reason, and Beatrice Revelation or Theology, we may perhaps suppose that Statius typifies something intermediate; such as Human Reason, generally enlightened by Christianity, but not specially instructed or interested therein; the cultivated lay' mind (not even the 'pious layman') in an age that has received the general impress of Christianity; a mind by which it is accepted and assumed rather than warmly embraced; one that is uncon

*

This seems clearly pointed out, inter alia, in Purg, xviii. 46-48 (Virgil loquitur):—

'Quanto ragion qui vede

Dirti poss' io, da indi in là t' aspetta
Pure a Beatrice, che opera è di fede.'

sciously rather than consciously under its influence. Christianity has, of course, lifted the minds, the ideas, the knowledge of mankind to a different level, though they may not be aware of what they precisely owe to it. They breathe a different atmosphere, though they may be unconscious of its ingredients and unaware of the degree or manner in which it has been changed for the better. Thus, as compared with any one who lived before Christianity was in the air,' the 'least in the kingdom of heaven is greater than he,' just as, since the establishment of the Copernican system, the veriest tyro in astronomy is enabled to start from a higher platform than would be attainable even by a Ptolemaic Newton. We only throw this out as a possible suggestion. Dante may have intended to create a type of this intermediate condition between Virgil and Beatrice, between the highest type of pre-Christian intellect, or merely human reason, and the fullest development of the soul enlightened by the treasures of wisdom and knowledge which are imparted by revelation and dogmatic theology.

Before leaving the subject of Dante and Statius we may call attention to one of the many passages cited or imitated by Dante from his works, for the sake of the interesting comment made upon it by Dante, and the consequent probability that it has served as the model for a celebrated passage in his own poem. In Conv. iv. 25, Dante gives as an illustration of Verecundia' the answer of Polynices to Adrastus (according to Statius, Theb. i. 668 seq.), when asked by him as to his parentage. His reply is:

'Cadmus origo patrum, tellus Mavortia Thebe,
Et genetrix Jocasta mihi' (1. 680).

Dante observes that through shame Polynices avoids mentioning the name of his father Edipus, but declares his ancestors, his country, and his mother. Dante having been thus struck by the tact and delicacy of Statius in this passage, and having himself drawn this inference from his language, we can scarcely doubt that we have an imitation of it in the exquisite episode of Manfred in Purg. iii., where Dante makes Manfred declare

'io son Manfredi,

Nepote di Costanza Imperadrice' (l. 112, 113).

He mentions his grandmother, but omits to name his mother (Bianca Lancia), since he was illegitimate.

*

* Benvenuto da Imola, quaintly, as usual, but here, it must be

Though some of Dante's references to Cicero involve special points of interest, we shall limit ourselves to the consideration of one only, which affords the key to one of the most important principles involved in his classification of sins in the Inferno. The two fundamental principles may be said to be (1) The distinction between sins of incontinence and of vicious habit (incontinenza and malizia), and the more venial character of the former. This is, of course, derived directly from Aristotle, and is definitely asserted upon his authority, by Dante in Inf. xi. 11. 70-90. (2) The distinction between sins of violence and sins of fraud, under one or other of which heads all kinds of malizia fall.

'D' ogni malizia ch' odio in ciel acquista,

Ingiuria è il fine, ed ogni fin cotale

O con forza o con frode altrui contrista' (ib. 11. 22–24). The former occupy Circle 7, and the latter Circles 8 and 9 of the Inferno, each with numerous subdivisions. The source of this distinction is not expressly acknowledged, but it has been pointed out (in the first instance, we believe, by Dr. Witte) that it is beyond doubt derived directly from a passage in Cicero, De Off. i. 13: Cum autem duobus modis, id est, aut vi aut fraude fiat injuria; fraus quasi vulpeculæ, 'vis, leonis videtur; utrumque homini alienissimum, sed 'fraus odio digna majore.' It will be observed that of these three clauses, the first is almost translated in the passage quoted above. The second is reproduced in the mouth of Guido da Montefeltro (Inf. xxvii. 74, 75) :'1' opere mie

Non furon leonine ma di volpe;'

and the sentiment of the third is implied in the next terzina to the one above quoted, in Canto xi.

'Ma perche frode è dell' uom proprio male,

Più spiace a Dio, e però stan di sutto
Gli frodolenti, e più dolor gli assale.'

It has also been pointed out in reference to this question -as to which before Dr. Witte's explanation much confusion and misapprehension prevailed, nor has it yet wholly disappeared from commentaries on the 'Inferno")-that Dante's familiarity with this portion of the De Officiis is shown by the fact that chapters xi.-xvii. are quoted or referred to by him in his several works no less than eight times.

admitted, rather flippantly, comments thus: 'He is like the mule who said he was the grandson of a horse, though he was the son of an ass!'

But little space remains for pointing out some of the practical purposes that may be served by a systematic study of the passages which Dante quotes or refers to in such authors as we have been considering. The first and most obvious arise from the bearing of such passages on (1) questions of textual criticism, (2) questions of interpretation.

As to the former, there is, unfortunately, a good deal of uncertainty about the text of all of Dante's works, the autographs of the whole of them having very early disappeared. In the case of the Divina Commedia, though an enormous num. ber of manuscripts (about 600) is known to exist, the earliest of them does not come within several years of Dante's death. On the other hand, five early commentaries have come down which were written within twelve years of his death, and it is curious to find various readings occasionally discussed even in some of them. Moreover, there are some cases in which readings which have very little support in manuscripts now existing are supported almost unanimously by the early commentators. This (we may remark in passing) illustrates the importance of admitting the evidence of patristic quotations to check the evidence of the surviving manuscripts of the New Testament. The chief difficulties, however, occur in the case of the Convito, the text of which is deplorably corrupt. Several readings have already been corrected, some with certainty, others with considerable probability, by the help of some of the very numerous direct quotations which abound in this work.

It is obvious that this is a very dangerous tool, and one that needs to be employed with the greatest precaution, and only by skilful hands. It would be difficult to imagine any process more uncritical than an offhand alteration of the text of an author simply to secure the accuracy of his quotations. It would assume on his part either an infallible memory, or else an habitual practice of verifying his references. The latter would be a practical impossibility before the days of printing, and it is none too common even in modern authors, with all their facilities for doing it. It would also involve another assumption, sometimes overlooked, viz. that the text which he used was the same as that now current. We cannot pursue this subject further than by merely giving two illustrations, one of the need of this last precaution, and the other of a safe and certain rectification of text by the help of a quotation. In the following passage from Purg. xxxiii. 49, the word Naiade is most inappropriate,

(inviting an obvious emendation), and yet is certainly the correct reading :

'Ma tosto fien li fatti le Naiade

Che solveranno questo enigma forte.'

How came Dante, it may be asked, to assign to the Naiads the astonishing function of solving riddles? Any one who refers to a modern edition of Ovid (Metam. vii. 759 seq.) will recognise the passage which Dante is imitating in this line, and, further, in this context generally. But he will also find that it reads thus :

Carmina Laiades non intellecta priorum
Solverat ingeniis.'

The path, and almost the duty, of conjectural emendation seems clearly pointed out. But on further inquiry we find that the reading Laiades is due to a manuscript discovered long since Dante's time, and that in his day the text of Ovid in this place was

Carmina Naiades non intellecta priorum
Solvunt ingeniis,'

The received text in Dante is therefore undoubtedly correct. The other illustration is this. In Conv. iv. 15, Dante quotes in an accurate Italian version a long passage (and therefore worth verification, and probably verified) of six consecutive lines from Ovid, Metam. i. 78-83. All goes well till we come to the words 'recens tellus seductaque nuper ab alto Æthere,' and then we find in all manuscripts the surprising reading, la recente terra di poco dipartita dal nobile corpo sottile e *diafano. It is evident that a marginal gloss has displaced the word 'etera' which puzzled some copyist, and the passage from Ovid enables us with confidence to restore the word. Apparently the copyist knew less about the 'ether' even than modern philosophers (auct. Lord Salisbury at the British Association of 1894); for it is most singular that in another passage Conv. ii. 15) where Dante is quoting another long passage in which the word 'æthera' occurs from Pror, viii. 27-30, all corresponds with the original till we come to 'Quando æthera firmabat sursum et librabat fontes aquarum,' and then the text of the manuscripts runs, quando suso fermava e sospendea le fonti dell' acque.' The pussling word is simply omitted altogether, since the senfence will construe in a fashion without it!

In employing quotations for the interpretation of the places which they occur we are generally on safer ground,