[7] In the Atlantic Ocean, over against the Pillars of Hercules, lay an island larger than Asia and Africa taken together, and in its vicinity were other islands. The ocean in which these islands were situated was surrounded on every side by main-land; and the Mediterranean, compared with it, resembled a mere harbor or narrow entrance. Nine thousand years before the time of Plato this island of Atlantis was both thickly settled and very powerful. Its sway extended over Africa as far as Egypt, and over Europe as far as the Tyrrhenian Sea. The further progress of its conquests, however, was checked by the Athenians, who, partly with the other Greeks, partly by themselves, succeeded in defeating these powerful invaders, the natives of Atlantis. After this a violent earthquake, which lasted for the space of a day and a night, and was accompanied with inundations of the sea, caused the islands to sink; and for a long period subsequent to this, the sea in that quarter was impassable by reason of the slime and shoals.—Plato, Tim., 24-29, 296; Crit., 108-110, 39, 43. The learned Gessner is of opinion that the Isle of Ceres, spoken of in a poem of very high antiquity, attributed to Orpheus, was a fragment of Atlantis. Kircher, in his "Mundus Subterraneus," and Beckman, in his "History of Islands," suppose the Atlantis to have been an island extending from the Canaries to the Azores; that it was really ingulfed in one of the convulsions of the globe, and that those small islands are mere fragments of it. Gosselin, in his able research into the voyages of the ancients, supposes the Atlantis of Plato to have been nothing more nor less than one of the nearest of the Canaries, viz, Fortaventura or Lancerote. Carli and many others find America in the Atlantis, and adduce many plausible arguments in support of their assertion.—Carli, Letters Amer.; Fr. transl., ii., 180. M. Bailly, in his "Letters sur l'Atlantide de Platon," maintains the existence of the Atlantides, and their island Atlantis, by the authorities of Homer, Sanchoniathon, and Diodorus Siculus, in addition to that of Plato. Manheim maintains very strenuously that Plato's Atlantis is Sweden and Norway. M. Bailly, after citing many ancient testimonies, which concur in placing this famous isle in the north, quotes that of Plutarch, who confirms these testimonies by a circumstantial description of the Isle of Ogygia, or the Atlantis, which he represents as situated in the north of Europe. The following is the theory of Buffon: after citing the passage relating to the Atlantis, from Plato's "Timæus," he adds, "This ancient tradition is not devoid of probability. The lands swallowed up by the waters were, perhaps, those which united Ireland to the Azores, and the Azores to the Continent of America; for in Ireland there are the same fossils, the same shells, and the same sea bodies as appear in America, and some of them are found in no other part of Europe."—Buffon's Nat. Hist., by Smellie, vol. i., p. 507.
[8] The first authentic description of the Mar di Sargasso of Aristotle is due to Columbus. It spreads out between the nineteenth and thirty-fourth degrees of north latitude. Its chief axis lies about seven degrees to the westward of the Island of Corvo. The smaller bank, on the other hand, lies between the Bermudas and Bahamas. The winds and partial currents in different years slightly affect the position and extent of these Atlantic "sea-weed meadows." No other sea in either hemisphere displays a similar extent of surface covered by plants collected in this way. These meadows of the ocean present the wonderful spectacle of a collection of plants covering a space nearly seven times as large as France.—Humboldt's Cosmos.
[9] See Appendix, No. II. (vol. II.)
[10] See Aristotle, De Mirab. Auscult., cap. lxxxiv., 84, p. 836, Bekk. This work, "A Collection of Wonderful Narratives," is attributed to Aristotle; the real compiler is unknown. According to Humboldt, it seems to have been written before the first Punic war.—Diodorus of Sicily, vol. xix. Aristotle attributes the discovery of the island to the Carthaginians; Diodorus to the Phœnicians. The occurrence is said to have taken place in the earliest times of the Tyrrhenian dominion of the sea, during the contest between the Tyrrhenian Pelasgi and the Phœnicians. The Island of the Seven Cities (see Appendix, No. II. (vol. II.)) was identified with the island mentioned by Aristotle as having been discovered by the Carthaginians, and was inserted in the early maps under the name of Antilla. Paul Toscanelli, the celebrated physician of Florence, thus writes to Columbus: "From the Island of Antilia, which you call the Seven Cities, and of which you have some knowledge," &c. In the Middle Ages conjectures were religiously inscribed upon the maps, as is proved by Antilia, St. Borondon (see Appendix (vol. II.)), the Hand of Satan, Green Island, Maida Island, and the exact form of vast southern regions. Humboldt refers the name of Antilia so far back as the fourteenth century. The earliest date given by Ferdinand Columbus is 1436. "Beyond the Azores, but at no great distance toward the west, occurs the Ysola de Antilia, which we may conclude, even allowing the date of the map to be genuine (in the library of St. Mark, at Venice, date 1436), to be a mere gratuitous or theoretic supposition, and to have received that strange name because the obvious and natural idea of antipodes has been anathematized by Catholic ignorance." He elsewhere says that "some Portuguese cosmographers have inserted the island described by Aristotle in maps under the name of Antilia."—Hist. of the Discovery of America, by Don Ferdinand Columbus, in Ker, vol. iii., p. 3-29.
The origin of the name Antilla, or Antilia, is still a matter of conjecture. Humboldt attributes to a "littérateur distingué" the solution of the enigma, from a passage in Aristotle's "De Mundo," which speaks of the probable existence of unknown lands opposite to the mass of continents which we inhabit. These countries, be they small or great, whose shores are opposed to ours, were marked out by the word porthornoi, which in the Middle Ages was translated by antinsulæ. Humboldt says that this translation is totally incorrect; however, the idea of the "littérateur distingué" is evidently the same as Ferdinand Columbus's. The following is the hypothesis favored by Humboldt: "Peut-être même le nom d'Antilia qui paraît pour la première fois sur une carte Vénitienne de 1436 n'est il qu'une forme Portuguaise donnée à un nom géographique des Arabes. L'étymologie que hasarde M. Buace me paraît très ingénieuse.... La syllabe initiale me paraît la corruption de l'article Arabe. D'al Tinnin et d'Al tin on aura fait peu à peu Antinna et Antilla, comme par un déplacement analogue de consonnes, les Espagnols ont fait de crocodilo, corcodilo et cocodrilo. Le Dragon est al Tin, et l'Antilia est peut-être, l'île des dragons marins."—Humboldt's Ex. Crit., vol. ii., 211.
Oviedo applies the relation of Aristotle to the Hesperian Islands, and asserts that they were the "India" discovered by Columbus. "Perchè egli (Colombo) conobbe come era in effetto che queste terre che egli ben ritrovava scritte, erano del tutto uscite dalla memoria degli uomin; e io per me non dubito che si sapissero, e possedessero anticamente dalli Rè de Spagna: e voglio qui dire quello che Aristotele in questo caso ne scrisse, &c.... io tengo che queste Indie siano quelle autiche e famose Isole Hesperide cosè dette da Hespero 12 Re di Spagna. Or come la Spagna e l'Italia tolsero il nome da Hespero 12 Re di Spagna cosi anco da questo istesso ex torsero queste isole Hesperidi, che noi diciamo, onde senza alcun dubbio si de tenere, che in quel tempe questo isole sotto la signoria della Spagna stessero, e sotto un medesmo Re, che fu (come Beroso dice) 1658 anni prima che il nostro Salvatore nascesse. E perchè al presente siamo nel 1535 della salute nostra, ne segue che siano ora tre milo e cento novantatre anni che la Spagna e'l suo Re Hespero signoreggiavano queste Indie o Isole Hesperidi. E come cosa sua par che abbia la divina giustizia voluto ritornargliele."—Hist. Gen. dell' Indie de Gonzalo Fernando d'Oviedo, in Ramusio, tom. iii., p. 80.
[11] "It is very possible that in the same temperate zone, and almost in the same latitude as Thinæ (or Athens?), where it crosses the Atlantic Ocean, there are inhabited worlds, distinct from that in which we dwell."[12]—Strabo, lib. i., p. 65, and lib. ii., p. 118. It is surprising that this expression never attracted the attention of the Spanish authors, who, in the beginning of the sixteenth century, were searching every where in classical literature with the expectation of finding some traces of acquaintance with the New World.
[12] "The idea of such a locality in a continuation of the long axis of the Mediterranean was connected with a grand view of the earth by Eratosthenes (generally and extensively known among the ancients), according to which the entire ancient continent, in its widest expanse from west to east, in the parallel of about thirty-six degrees, presents an almost unbroken line of elevation."—Humboldt's Cosmos.
[13] "D'Anville a dit avec esprit que la plus grande des erreurs dans la géographie de Ptolémée a conduit les hommes à la plus grande découverte de terres nouvelles c'est, à dire la supposition que l'Asie s'étendait vers l'est, au delà du 180 degré de longitude."
Both Strabo and Aristotle speak of "the same sea bathing opposite shores," Strabo, lib. i., p. 103; lib. ii., p. 162. Aristotle, De Cælo, lib. ii., cap. 14, p. 297. The possibility of navigating from the extremity of Europe to the eastern shores of Asia is clearly asserted by the Stagirite, and in the two celebrated passages of Strabo. Aristotle does not suppose the distance to be very great, and draws an ingenious argument in favor of his supposition from the geography of animals. Strabo sees no obstacle to passing from Iberia to India, except the immense extent of the Atlantic Ocean. It is to be remembered that Strabo, as well as Eratosthenes, extend the appellation of Atlantic Sea to every part of the ocean.—Humboldt's Géog. du Nouveau Continent.