XL] THE 'ATLANTIC ISLANDS.1 225
In this connexion we may notice a discovery that forcibly
impressed the men of that time, the mention of The'Atian
which occurs in the course of Sertorius' career in tic islands*
Spain. In the year 81 B.C. that adventurous com- tMadeira)t
mander, when he found himself unable to make head against the
forces which Sulla had sent to oppose him in that country,
happened to meet near the mouth of the Baetis with some seamen
who had recently visited the 'Atlantic Islands/ These they
described as being two in number, separated by a very narrow
channel, and lying in the open sea at a distance of ten thousand
stadia (1,000 geographical miles) from the African coast. The
climate of the islands they reported to be delightfully temperate,
exempt from cold and violent winds and from excessive rain, with
a soft and moist air, which not only rendered the soil fertile for
cultivation, but produced self-sown fruits in great abundance. The
account thus given took such hold on the imagination of Sertorius,
that he was seized with a strong desire to betake himself to this
spot, where he might " escape from tyranny and unceasing wars,
and live in tranquillity"; but he was forced to desist from the
project by the unwillingness of the Cilician pirates who formed the
crews of his ships to accompany him. As was natural, these
islands were identified with the ' Islands of the Blessed/ which had
been celebrated from early days in Greek poetry—" where is no
snow, nor yet great storm, nor any rain; but alway ocean sendeth
forth the breeze of the shrill West to blow cool on men1" : indeed,
we are told that the barbarians themselves believed that in them
were to be found the Elysian plains and the Abodes of the Happy,
of which Homer had sung*. Though the distance from the
continent which is attributed to these islands must in any case
have been a great exaggeration, yet it seems impossible to
regard the Canaries, lying as they do within easy reach of the
'African coast, as corresponding to them; and the circumstance
that they are spoken of as two only, suggests that Madeira
and the neighbouring Porto Santo were meant, rather than such
groups as the Azores or the Cape Verde islands. The humidity
and equable character of the climate of Madeira, also, and the
great productiveness of the ground, are in favour of this view.
1 Horn. Od., 4. 566—8. . * Plutarch, Sertor^ 8, $.
T. 15