noticed inf. under Aristotle, Nos. 11 and 23. I have not had the opportunity of applying such a test to some of the disputed Epistles, but as there again we depend mainly on internal evidence, I think it is probable that the statistics of quotations here collected might be profitably employed. In these cases it is to be remembered that we must be content to catch at straws,' since all the available arguments are of a slender kind, but if several agree in pointing in the same direction, their contributory force is not to be neglected.

But there is no doubt that it is in the light thrown upon interpretation that the chief use of such references as these is to be sought. I have often thought that the most generally useful commentary on the Divina Commedia in existence is the invaluable Concordance of Dr. Fay. It best of all enables us to give effect to the cardinal principle insisted on by Giuliani, 'Dante spiegare con Dante.' And it is scarcely less important to interpret him also by the help of those authors whose language and sentiments he is consciously reproducing from time to time by quotation, imitation, or allusion. This enables us to place his own thoughts in their proper and original setting, and to see them as they shaped themselves in his own mind.

It remains to explain the rough classification which I have attempted of the quotations registered under three heads indicated by the letters a, b, and c, intending thereby to distinguish three different degrees (so to speak) of quotation.

(a) Cases of direct quotation, either (as in most cases) acknowledged as such with a reference to the author or work quoted, or else so obviously intended to be distinct quotations that a modern writer or editor would undoubtedly insert the usual marks of quotation.

(b) Cases where the quotation is not verbally acknowledged, but where the reference to a definite passage of some earlier author is almost equally certain or unmistakable. Under this head are included cases where a passage in Dante has evidently been suggested, either in form or matter, by some Scriptural or classical original, as is constantly the case in respect of historical or mythological details. The source of

these can often be as confidently identified as if there were a verbal quotation, as for instance the reference to Elisha and the fiery chariot of Elijah in Inf. xxvi. 34 seqq., or many stories from the Metamorphoses of Ovid in the Inferno and Purgatorio. (c) This last class includes some cases where the reference is in various degrees probable, but not altogether beyond doubt; and also others where Dante's language is rather that of imitation, adaptation, or allusion, than of actual quotation, though such allusion may sometimes admit of no doubt. But I have always (as I have stated above) avoided including mere 'illustrations,' mere coincidences of thought or expression, unless (in my judgement at any rate) there was reason to suppose that Dante's language was moulded or modified by a reminiscence of the passage to which I have referred. I have endeavoured in doubtful cases to apply this test :-Would Dante's language or thought have been likely to take this precise form, but for the influence exercised on them by the recollections of such and such a passage in Scripture or some classical author? Thus, for example, I have ventured to include a reference to the Laodiceans of Rev. iii. 16 as almost certainly suggesting the idea of Inf. iii. 34 seqq., though there is no formal reference or quotation.

I do not attach much importance to this classification, since opinions would often differ as to the proper letter to be assigned, and I have often hesitated myself. But I have thought it might sometimes be of use, and it can at any rate do no harm.

In the great majority of instances the parallel passages are simply registered in the Index, and may be left to speak for themselves. In the case of those selected for discussion some point of interest seemed to be involved, such as the bearing on the form or the interpretation of the text of Dante, or various other matters to which it seemed worth while to call attention from time to time. The passages so discussed are in the first Index arranged under the authors from whom the quotations are taken, with a few general preliminary observations when necessary as to Dante's use of the author in question.

Under each author different principles of grouping have been adopted as seemed most convenient in each case, so as to keep together passages having some common point of connexion or of interest. The order, however, is not of much consequence, since references are given in the Index to the pages where each passage will be found.

I now proceed to speak separately of Dante's treatment of the principal authors from whom he quotes, and to discuss the passages which seem to me to call for some special comment under each author. We commence with

THE VULGATE.

It is important to remember that in this version only were the writings either of the Old or New Testament known to Dante, because the references occasionally cease to be appropriate at all when compared with the reading or rendering of our Authorized Version 1. Also the division of the verses, especially in the Apocrypha and sometimes in the Psalms, is different. I have already observed that Dante's language is full of Scriptural phraseology, even when definite quotation is not in question. And it is scarcely necessary to add that his reverence for the authority of Scripture is unbounded, and any statement therein found, either directly, or as transformed by the strange processes of allegorical interpretation and application then universally recognized as valid 2, is for him an absolute final and self-sufficient warrant for any point with which it deals. The following passages out of many others will be sufficient to illustrate this: -Par. v. 76; xix. 82-4; xxiv. 91-102, 133-8; Conv. II. ix. ll. 113 seqq.; De Mon. III. iv. ll. 81-91, &c.

I will first take together a few quotations which seem to be

These are some instances among the passages discussed or indexed:Inf. xxxi. 10; Purg. x. 58; xi. 116; xvii. 26; Par. xi. 53; xxix. 28–30; Conv. II. vi. 1. 34, &c. Also one of the stock passages by which the central position of Jerusalem was proved by mediaeval writers-'Deus. . . operatus est salutem in medio terrae,' Ps. lxxiii. 12-is quite different in E. V. See Ps. lxxiv. 13. See Conv. II. i.; Conv. IV. xvii. ll. 85-111; xxii. ll. 145 seqq.; De Mon. I. xv. ll. 19-24; II. xviii.; III. ix. ad fin. &c.

available for purposes of textual criticism, then some that bear upon the interpretation of the passages in which they occur, and afterwards those which offer other various points of interest. In a very few cases no reference will be found in the index to a passage that is discussed. This is either because the tendency of the discussion is to show that the reference is wrongly assumed, or else because, though some reference is implied, it is too general to be entered against any one particular passage in the original author.

Inf. i. 61.]

(1)

Mentre ch' io rovinava in basso loco.

Here we have the various readings rovinava (or ruinava), ritornava, and rimirava. Apart from other considerations, such as the greater vigour and picturesqueness of the first of these words, and its use again in allusion to this same scene in Par. xxxii. 138, and to a similar one in Inf. xx. 35, it is also strongly supported by a reference to Prov. iv. 19, 'Via impiorum tenebrosa, nesciunt ubi corruunt.' This passage is directly quoted in Conv. IV. vii. Il. 98 seqq., and it being of course translated, the last words appear as 'non sanno dove rovinano,' 'They know not at what they stumble' (E. V.). The word then here means, 'While I was blundering or stumbling in the valley.' The contrast of the joyous and bright mountain in the previous lines, and again in ll. 77, 78, may well be compared with preceding verse in Proverbs, 'Justorum autem semita quasi lux splendens procedit, et crescit usque ad perfectam diem.' Thus, while this cannot be registered exactly as a 'quotation,' yet the comparison together of the several passages which I have given seems fairly to throw light on the question of the reading.

Purg. vi. 109.]

(2)

Vedi la pressura
De' tuoi gentili.

The reading pressura is much better supported than oppressura, which is both a very obvious word in itself and also may easily have arisen from a misunderstanding of lapressura.

But besides, the reading pressura is made almost certain by its use exactly in this sense in St. Luke xxi. 25, 'pressura gentium,' 'distress of nations.' (Even the word gentium seems echoed in gentili.) The word occurs also in St. John xvi. 33, In mundo pressuram habebitis.' Also in De Mon. III. ix. 1. 48, Dante uses the word again in evident allusion to these passages. A quotation of one of our Lord's last warnings to the disciples is explained, 'hoc etiam dicebat praemonens eos pressuram futuram.'

(3)

Purg. x. 120. Già scorger puoi come ciascun si picchia.

Here not only is there another reading nicchia (less well supported), but also commentators give several different explanations of picchia (see Scart. note h. .). But both the reading and interpretation seem to be fixed by the obvious reference to the publican in St. Luke xviii. 13, 'Publicanus

nolebat nec oculos ad coelum levare sed percutiebat pectus suum,' &c. His attitude is reproduced in 1. 116 (a terra li rannicchia), and his action in the words above quoted. The propriety of this is further enhanced when we remember that it is pride which is here being expiated by all the conditions of the most profound humility.

(4)

Purg. xiv. 133. Anciderammi qualunque m'apprende.

This evidently represents Gen. iv. 14, 'Omnis igitur qui invenerit me, occidet me.' We cannot therefore doubt that the reading (also better supported) m' apprende is to be preferred to mi prende, which would convey quite a different meaning.

(5)

Purg. xxiv. 126. Quando ver Madian discese i colli.

The reading distese is so clearly a mere copyist's blunder for discese that it is hardly worth while (except as showing Dante's graphic recollection of the original) to notice that the latter is evidently required by the passage in Judges which is referred to. In chap. vii. 8, 9, we read 'Castra autem Madian erant subter in valle. Eadem nocte dixit

E