being that the souls of men arrive at the Moon, wherein lie the
Elysian Fields of Homer.

The Greek historian, Diodorus Siculus, has a similar account with
curious details of an "island" which might very well have been part of a
continent. Columbus believed to the last that Cuba was a continent.

In the ocean, at the distance of several days' sailing to the west,
there lies an island watered by several navigable rivers. Its soil
is fertile, hilly, and of great beauty.... There are country
houses handsomely constructed, with summer-houses and flower-beds.
The hilly district is covered with dense woods and fruit-trees of
every kind. The inhabitants spend much time in hunting and thus
procure excellent food. They have naturally a good supply of fish,
their shores being washed by the ocean.... In a word this island
seems a happy home for gods rather than for men (v. 19).

Another Greek writer, Lucian, in one of his witty dialogues, refers to
an island in the Atlantic, that lies eighty days' sail westward of the
Pillars of Hercules--the extreme limit of the ancient world, as has
already been seen. Readers of Henry Fielding and admirers of Squire
Westers will remember how in the London of the eighteenth century the
limits of Piccadilly westward was a tavern at Hyde Park corner called
the _Hercules' Pillars_, on the site of the future Apsley House.[1]

Although neither Greek nor Roman navigators were likely to attempt a
voyage into the ocean beyond the Straits of Gibraltar, yet a trading
vessel from Carthage or Phenicia might easily have been driven by an
easterly gale into, or even across, the Atlantic. Some involuntary
discoveries were no doubt due to this chance, and the reports brought to
Europe were probably the germs of such tales as the poets invented about
the fair regions of the West. In Celtic literature, moreover, "Avalon"
was placed far under the setting sun beyond the ocean--Avalon or
"Glas-Inis" being to the bards the Land of the Dead, marvelous and
mysterious.

[Footnote 1: Tom Jones, xvi. chap. 2, 3, etc.]

In English literature of the middle ages there is a remarkable passage
relating to our present subject, which was written long before that rise
of the New Learning mentioned at the beginning of this chapter. It is a
statement made by Roger Bacon, the greatest of Oxonian scholars of the
thirteenth century, who, long before the Renascence, did much to re