religious, not in any way scientific-its true object is to give expression to man's sense of responsibility, which it does in the form of a vivid history, or spectacle, of the connected lifestages of an immortal personality. This moral and religious object, however, is served best, if the history or spectacle, though carefully presented as a creation of fancy, is not made too fantastical, but is kept at least consistent with "modern science." 1 It is of the greatest importance that the student of the philosophy of Plato's Myths should learn to appreciate the terms of this alliance between Myth and Science; 2 and I do not know how the lesson can be better learnt than from parallel study of Dante's Divina Commedia, in which all the science-moral and physical-of the age is used to give verisimilitude to the great pos of medieval Christianity. Fortunately, no better instances of the art with which Dante presses Science into the service of Myth could be found than in his treatment of a subject which has special interest for us here, in connection with the geography and geology of the Phaedo Myth. This brings me to the second head of observations which I have to offer on the Phaedo Myth.

II

In this section I wish to draw attention to the parallel between Plato's geography of Tartarus and the True Surface of the Earth, and Dante's geography of Hell and the Mount of Purgatory with the Earthly Paradise on its summit.

The parallel is close. On the one hand, the Phaedo Myth and the Divina Commedia stand entirely alone, so far as I know, among Eschatological Myths in making Tartarus or Hell a chasm bored right through the globe of the Earth (διαμπερὲς τετρημένον δι ̓ ὅλης τῆς γῆς, Phaedo, 111 E; Inferno, xxxiv. sub fin.), with two antipodally placed openings. On the other hand, while the Phaedo Myth stands alone among Plato's Eschatological Myths in describing a lofty terrestrial region raised, above the elements of water and air, up into the

1 Aristotle's canon applies-προαιρεῖσθαί τε δεῖ ἀδύνατα εἰκότα μᾶλλον ἢ δυνατὰ ἀπίθανα.—Poet. 1460 a 30.

2 In this connection the reader should turn to Prof. Dill's illuminating remarks on the mixture of science with devotional allegory and myth in the Commentary of Macrobius on Cicero's Dream of Scipio: Roman Society in the Last Century of the Western Empire, Book i. ch. iv. pp. 88-90, ed. 1.

element of fire or aether, Dante also, in agreement with a common medieval belief, places the Earthly Paradise on the top of a mountain-his own Mount of Purgatory—which rises up into the element of fire.

1

The "Earthly Paradise" of the Phaedo Myth probably owes a good deal to the Homeric Olympus; and the Earthly Paradise of medieval belief and of the Divina Commedia may have derived at least its altitude from the same source. But the description of Tartarus as bored right through the Earth, unique in Greek mythology, in no way countenanced by Virgil, and yet reappearing in the Inferno, which is so largely modelled on the Sixth Book of the Aeneid-this is surely a strange coincidence. The Timaeus (in the version of Chalcidius) was, it would appear, the only work of Plato which Dante knew directly. There is no evidence whatever-unless this coincidence be regarded as evidence that he was acquainted with the Latin version of the Phaedo which was made in the twelfth century. It is possible, however, but I hardly think likely, that the passage in the Meteorologica (ii. 2, 355 b, 32 ff.), in which the Phaedo description of Tartarus is referred to, may have given Dante the idea of an antipodal exit from Hell; although it is to be noted that Aristotle, in criticising the hydrostatics of the Phaedo Myth, curiously enough omits to quote, or paraphrase, Plato's emphatic StaμTEρès тEтρNμévov; and S. Thomas does not make good the omission in his commentary on the Aristotelian passage. I do not think that any one reading the Aristotelian passage, without having read the Phaedo, would easily gather that the Tartarus of the Phaedo is bored right through the Earth. Aristotle is concerned to show that the theory of a central aiúpa, or oscillation, gives a wrong explanation of the origin of seas and rivers; and, more suo, he is careless in his description of the theory to which he objects. Although the hydrostatics of the Quaestio de Aqua et Terra3 agree in the main with 1 See Moore's Studies in Dante, first series, p. 156, and Toynbee's Dante Dictionary, s.v. "Platone.'

2 See Rashdall's Universities of Europe in the Middle Ages, i. 37, ii. 744, and Immisch, Philologische Studien zu Plato, pp. 33, 34. Henricus Aristippus (Archdeacon of Catania) translated the Phaedo and Meno in 1156. There is a MS. of his translation in Corpus Christi College, Oxford (243), written in 1423; see Coxe, ii. 100.

3 With regard to the authenticity of this treatise see Moore's Studies in Dante, second series, pp. 303 ff.

those of the Meteorologica, the Inferno is not influenced by the Meteorologica. The Inferno follows the traditional mythology in supposing subterranean rivers, and, indeed, agrees with the account of these rivers given in the Phaedo, to the extent, at least, of regarding them as forming a single system of waters connected somehow with waters on the surface of the Earth. Dante may have been helped to this view by Brunetto Latini, who speaks, very much in the same way as Plato does, of waters circulating in channels through the Earth, like blood through the veins of the body, and coming out in springs. But mark how the Poet uses these mere hydrostatics-how his genius transforms the physical relation between the living world and Tartarus into a moral relation! It is the tears of this world that flow in the rivers of Dante's Hell.2

1

Let me close this passage on Plato's Tartarus and Dante's Hell with the remark that an antipodal exit from Hell, near the Mount of Purgatory, is almost necessary to the movement of the Commedia. If such an exit-whether derived directly or indirectly from the Phaedo, or obtained from some other source-did not already exist among Dante's mythological data, he would practically have been obliged to invent it, and offer some explanation of it, such as that which he actually offers the Fall of Lucifer (Inf. xxxiv.).

Now to pass on to the parallel between Plato's "True Surface of the Earth" and Dante's Earthly Paradise on the top of the Mount of Purgatory:-Dante's Mount of Purgatory is definitely a part of this Earth. It is an island, antipodal to Jerusalem, in the middle of the ocean which covers the southern hemisphere. This island rises up, in a series of circular terraces, into one lofty height on which is situated the Earthly Paradise,-where our first parents were created,-where the souls which have been purified by 1 See Schmidt, über Dantes Stellung in der Geschichte der Kosmographie, I. Teil, de Aqua et Terra (1876), p. 7.

2 Inferno, xiv. Dante probably profited by the crude fancy of predecessors in the matter of the contents of the infernal rivers; see Cary on Inf. xii. It is perhaps worth noticing here that Dante's River of Blood (Inf. xii.) has its parallel in the Scottish ballad of Thomas the Rhymer :

It was mirk mirk night and there was nae stern-light,
And they waded through red bluid to the knee;

For a' the bluid that's shed on earth

Rins through the springs o' that countrie (i.e. Elf-land).

penance during their ascent of the Mount are gathered together, before they drink the waters of Lethe and Eunoè, the twin streams of this Paradise, and are translated into the

Heavenly Paradise. That Purgatory is a real place, on the surface of this globe, which an adventurous voyager from our hemisphere might possibly reach vnì peλaívy, is suggested with consummate art in the Inferno, Canto xxvi., where Ulysses describes his last voyage-how, with Ceuta on his left and Seville on his right, he sailed out through the Straits, and south over the ocean for five months, till the stars of the northern hemisphere sank beneath the horizon, and new stars appeared in the sky, and he sighted

A Mountain dim, loftiest, methought,

Of all I e'er beheld 1

and then the storm burst which overwhelmed him.

3

Dante's Mount of Purgatory-for that was the land which Ulysses sighted is identical with the lofty mountain on the top of which medieval belief placed the Earthly Paradise; but Dante apparently drew entirely on his own imagination when he localised Purgatory on its slopes. This Mountain of the Earthly Paradise rises, according to the medieval belief, as high as the Lunar Sphere -i.e. its upper parts are above the air, in the aether or fire, like Plato's True Surface of the Earth. Hence, as S. Thomas explains, the Earthly Paradise was not reached by the flood. S. Thomas further remarks that Enoch and Elias are said to be now in it; also, that it is said to be sub aequinoctiali circulo; but he will not vouch for its exact position, only expressing his belief that it must be in a " temperate clime." 5 The Arabians, whose geographical treatises, and epitomes of the Greek geographers, Dante knew in Latin versions, spoke of a great

1 Cary's translation.

Purga-
It

2 See Scartazzini (Companion to Dante, Butler's Transl. p. 419). tory, so far as form and position go, is a creation quite of the poet's own." may, I think, have relationship to the "steep hill of virtue" which the Stoics climbed; see Lucian, Vera Hist. ii. 18-no Stoics were to be seen in the Fortunate Island, because they were climbing this hill: Tŵv dè Σtwikŵv ovdels παρῆν· ἔτι γὰρ ἐλέγοντο ἀναβαίνειν τὸν τῆς ἀρετῆς ὄρθιον λόφον.

See S. Thom. Aqui. Summa, i. 102, 2.

Cf. Schmidt, Cosmographie des Dante, p. 23.

5 Summa, i. 102, 2.

See Lelewel, Histoire de la Géographie, i. lxxxv., and Toynbee's Dante Dictionary, arts. Alfergano" and "Tolommeo1."

1

mountain in the far south. It is called Mons Caldicus by Albertus, and Mons Malcus by Roger Bacon, who places it in India. The view that this mountain, identified by the Christian Schoolmen with the seat of the Earthly Paradise, is an island antipodal to Jerusalem in the middle of the Southern Ocean (Purg. iv. 70), was due entirely, it would seem, to Dante's own "scientific imagination" or "mythopoeic faculty." According to the doctrine of Orosius, generally accepted in Dante's time, there is no land at all in the southern hemisphere. If there were land, its inhabitants would be cut off from those of the orbis notus—the unity and continuity of the human race, postulated by the command, "Go ye into all the world and preach the gospel to every creature," would not exist. The ideal of one Church and one Empire (and one Aristotelian Philosophy, as Dante adds in the Convivio, iv. 6) requires the geographical condition of one continuous οἰκουμένη. Dante's antipodal island, however, being peopled only by the souls of the departed, is in no way inconsistent with the teleological geography of Orosiusindeed, is made, with consummate art, to corroborate it; for the cause which produced the solitary island of Purgatory in

1 Meteor, ii. 2. 7. Cf. Schmidt, Cosm. d. Dante, p. 23. 2 Op. Maj. pp. 192, 195, ed. princ. Jebb, London.

3 See Scartazzini's Companion to Dante, p. 419, Butler's Eng. Transl. It is, however, an island in the Exeter Book (an Anthology of Anglo-Saxon Poetry given to the Library of Exeter Cathedral by Leofric, first Bishop of Exeter, 1050-1071): see Exeter Book, edited by Israel Gollancz for the Early English Text Society, 1895, poem on the "Phoenix," pp. 200 ff.: "The Earthly Paradise is in eastern parts. it is all plain. is an island. There the door of Heaven's Realm is oft-times opened. It is green and flowery. There is no rain there, nor snow nor frost nor fire. It is neither too hot nor too cold. The plain (which is quite smooth) is higher than any mountain by 12 fathom measures. It escaped the flood. . . . It shall abide perennially blooming till the Day of Judgment. Water falls not there, but rises from the turf in the midst of the forest each month of the year, and irrigates the grove [we are reminded of Dante's Lethe and Eunoè]. The beautiful grove is inhabited by the Phoenix "-which the Poet then goes on to describe.

It ought to be mentioned that Claudian (Idyll. i. 1. Phoenix) makes "the Earthly Paradise" an island :

Mr. Toynbee, however, thinks it doubtful whether Dante had any acquaintance with Claudian (see Dante Dict. art. "Claudianus "). Benvenuto da Imola, in his Commentary on the Divina Commedia, quotes Claudian several times, describing him, erroneously, as a Florentine; see Mr. Toynbee's Index of Authors quoted by Benv. da Imola in his Commentary on the D. C. (Annual Report of the Dante Society, Cambridge, Mass., 1901).

4 Ŏrosius, Hist. adv. paganos, i. 2, §§ 87-89; vi. 22, § 1; vii. 1; vii. 3, 4; and cf. Moore's Studies in Dante, first series, pp. 279 ff.