Legendary islands of the Atlantic; a study in medieval geography

[Autolocus]

Legendary islands of the Atlantic; a study in medieval geography

HANDBOUND
AT THE



UNIVERSITY OF
TORONTO PRESS




LEGENDARY ISLANDS
OF THE ATLANTIC



AMERICAN GEOGRAPHICAL SOCIETY

RESEARCH SERIES NO. 8

W. L. G. JOERG, Editor

LEGENDARY ISLANDS
OF THE ATLANTIC

A Study in Medieval Geography

BY
**

WILLIAM H: BABCOCK

Author of "Early Norse Visits to North America"





NEW YORK

AMERICAN GEOGRAPHICAL SOCIETY
1922



COPYRIGHT, 1922

BY

THE AMERICAN GEOGRAPHICAL SOCIETY
OF NEW YORK




THE CONDE NAST PRESS
GREENWICH, CONN.

[Autolocus]

CONTENTS

CHAPTER PAGE

I INTRODUCTION i

II ATLANTIS n

III ST. BRENDAN'S EXPLORATIONS AND ISLANDS ... 34

IV THE ISLAND OF BRAZIL 50

V THE ISLAND OF THE SEVEN CITIES 68

VI THE PROBLEM OF MAYDA 81

VII GREENLAND OR GREEN ISLAND 94

VIII MARKLAND, OTHERWISE NEWFOUNDLAND 114

IX ESTOTILAND AND THE OTHER ISLANDS OF ZENO . . 124

X ANTILLIA AND THE ANTILLES 144

XI CORVO, OUR NEAREST EUROPEAN NEIGHBOR . . . 164

XII THE SUNKEN LAND OF Buss AND OTHER PHANTOM

ISLANDS 174

XIII SUMMARY 187

INDEX 191

[Autolocus]

The following chapters are reprinted, with modifications, from th
Geographical Review: III, Vol. 8, 1919; V, Vol. 7, 1919; VI, Vol. 9,
1920; VIII, Vol. 4, 1917; X, Vol. 9, 1920; XI, Vol. 5, 1918.



LIST OF ILLUSTRATIONS

(All illustrations, except Figs, i, 15, and 23, are reproductions of

medieval maps. The source is indicated in a general way in each

title; the precise reference will be found in the text where the map is

first discussed.}

FIG. PAGE

1 Map of the Sargasso Sea, 1 72,000,000 28

2 The Pizigani, 1367 (two sections) 40-41

3 Beccario, 1426 facing 45

4 Dalorto, 1325 51

5 Catalan map, 1375 58

6 Nicolay, 1560 62

7 Catalan map, about 1480 64

8 World map in portolan atlas, about 1508 (Egerton

MS. 2803) facing 74

9 Desceliers, 1546 76

10 Ortelius, 1570 77

11 Ptolemy, 1513 82

12 Prunes, 1553 88

13 Coppo, 1528 97

14 Bishop Thorlaksson, 1606 98

15 Map of the early Norse Western and Eastern Settlements

of Greenland, 1 :6,4OO,ooo 103

16 Clavus, 1427 104

17 Donnus Nicolaus Germanus, after 1466 facing 105

1 8 Sigurdr Stefansson, 1590 107

19 Zeno, 1558 126

20 Beccario, 1435 152

21 Pareto, 1455 158

22 Benincasa, 1482 160

23 Representation of Corvo on fourteenth- and fifteenth-

century maps as compared with its present outline . . 172

24 Buss Island, probably 1673 i?6

25 Bianco, 1436 179

[Autolocus]

CHAPTER I
INTRODUCTION

We cannot tell at what early era the men of the eastern Medi-
terranean first ventured through the Strait of Gibraltar out on
the open ocean, nor even when they first allowed their fancies
free rein to follow the same path and picture islands in the great
western mystery. Probably both events came about not long
after these men developed enough proficiency in navigation to
reach the western limit of the Mediterranean. We are equally in
lack of positive knowledge as to what seafaring nation led the way.

The weight of authority favors the Phoenicians, but there
are some indications in the more archaic of the Greek myths
that the Hellenic or pre-Hellenic people of the Minoan period
were promptly in the field. These bequests of an olden time are
most efficiently exploited, in the matter-of-fact and very credulous
/'Historical Library" of Diodorus Siculus, 1 about the time of Julius
Caesar, who feels himself fully equipped with information as to
the far-ranging campaigns of Hercules, Perseus, and other wor-
thies. His identifications of tribes, persons, and places find an
echo which may be called modern in Hakluyt's map of I58y, 2
illustrating Peter Martyr, which shows the Cape Verde Islands
as Hesperides and Gorgades vel Medusiae. But this, though
curious, is, of course, irrelevant as corroboration. Diodorus
himself was a long way from his material in point of time, but
from him we may at least possibly catch some glimmer of the
origin of the mythical narratives, some refraction of the events
that suggested them.

l The Historical Library of Diodorus the Sicilian, in 15 Books, to which are
added the fragments of Diodorus, and those published by H. Valesius, I. Rhodo-
mannus, and F. Ursinus, transl. by G. Booth, Esq., 2 vols., London, 1814; reference
in Vol. i, Bk. 3, Ch. 4, p. IQS. and Bk. 4, Ch. i, pp. 235 and 243.

*A. E. Nordenskiold: Facsimile-Atlas to the Early History of Cartography,
transl. by J. A. Ekelof and C. R. Markham, Stockholm, 1889. p. 131.

[Autolocus]

2 INTRODUCTION

EARLY ACCOUNTS OF BIG SHIPS

Small coasting, and incidentally sea-ranging, vessels must be of
great antiquity, for the record of great ships capable of carrying
hundreds of men and prolonging their voyages for years extends
very far back indeed. We may recall the Scriptural item inci-
dentally given of the fleets of Hiram, King of Tyre, and Solomon,
King of Israel: "For the king had at sea a navy of Tharshish
with the navy of Hiram: once in three years came the navy of
Tharshish, bringing gold, and silver, ivory, and apes, and pea-
cocks." 3 Tharshish is generally understood to have been Tar-
tessus by the Guadalquivir beyond the western end of the Medi-
terranean. The elements of these exotic cargoes indicate, rather,
traffic across the eastern seas. No doubt "ship of Tarshish" had
come (like the term East Indiaman) to have a secondary meaning,
distinguishing, wherever used, a special type of great vessel of
ample capacity and equipment, named from the long voyage
westward to Spain, in which it was first conspicuously engaged.
But this would carry back we know not how many centuries the
era of huge ships sailing from Phoenicia toward the Atlantic and
seemingly able to go anywhere; with the certainty that lesser
craft had long anticipated them on the nearer laps of the journey
at least.

Corroboration is found in the utterances of a Chinese observer,
later in date but apparently dealing with a continuing size and
condition. "There is a great sea [the Mediterranean], and to the
west of this sea there are countless countries, but Mu-lan-p'i
[Mediterranean Spain] is the one country which is visited by the
big ships. . . Putting to sea from T'o-pan-ti [the Suez of to-
day] . . . after sailing due west for full an hundred days, one
reaches this country. A single one of these (big) ships of theirs
carries several thousand men, and on board they have stores of
wine and provisions, as well as weaving looms. If one speaks of
big ships, there are none so big at those of Mu-lan-p'i." 4

[Autolocus]

I Kings, 10: 22.

Chau Ju-Kua: His Work on Chinese and Arab Trade in the Twelfth and
Thirteenth Centuries Entitled Chu-fan-chi, transl. and annotated by Friedrich
Hirth and W. W. Rockhill, St. Petersburg. 1911, p. 142.



THE ATLANTIS LEGEND 3

This statement is credited to only a hundred years before
Marco Polo. One naturally suspects some exaggeration. But a
parallel account, nearly as expansive and very circumstantial, is
given in the same work concerning giant vessels sailing in the
opposite direction some six hundred years earlier. It begins:
"The ships that sail the Southern Sea and south of it are like
houses. When their sails are spread they are like great clouds in
the sky." Professor Holmes, drawing attention to these passages
(which he quotes), very justly observes, "who shall say that the
mastery of the sea known to have been attained in the Orient
500 A. D. had not been achieved long prior to that date?" 5

[Autolocus]

THE ATLANTIS LEGEND

We may be safe in styling Atlantis (Ch. II) the earliest mythi-
cal island of which we have any knowledge or suggestion, since
Plato's narrative, written more than 400 years before Christ, puts
the time of its destruction over 9,000 years earlier still. It seems
pretty certain that there never was any such mighty and splendid
island empire contending against Athens and later ruined by
earthquakes and engulfed by the ocean. Atlantis may fairly be
set down as a figment of dignified philosophic romance, owing its
birth partly to various legendary hints and reports of seismic and
volcanic action but much more to the glorious achievements of
Athens in the Persian War and the apparent need of explaining a
supposed shallow part of the Atlantic known to be obstructed
and now named the Sargasso Sea. Perhaps Plato never intended
that any one should take it as literally true, but his story undoubt-
edly influenced maritime expectations and legends during medi-
eval centuries. It cannot be said that any map unequivocally
shows Atlantis; but it may be that this is because Atlantis van-
ished once for all in the climax of the recital.

[Autolocus]

PHOENICIAN EXPLORATION

It may be that Phoenician exploration in Atlantic waters was
well developed before noo B.C., when the Phoenicians are

s W. H. Holmes: Handbook of Aboriginal American Antiquities, Bur. of Amer.
Ethnology, Bull. 60, Part I, Smithsonian Instn., Washington. D. C., 1919. P- 27-

[Autolocus]

4 INTRODUCTION

alleged to have founded Cadiz on the ocean front of southern
Spain; but its development at any rate could not have been
greatly retarded after that. The new city promptly grew into
one of the notable marts of the world, able during a long period
to fit out her own fleets and extend her commerce anywhere.
It is greatly to be regretted that we have no record of her dis-
coveries. Carthage, a younger but still ancient Tyrian colony,
farther from the scene of western action, was not less enterprising
and in time quite eclipsed her; but at last she fell utterly, as did
Tyre itself, whereas Cadiz, though no longer eminent, continues
to exist. However, in her prime Carthage ranged the seas pretty
widely; according to Diodorus Siculus, she was much at home
in Madeira, 6 and her coins have been found off the shore of
distant Corvo of the Azores. But it cannot be said that any of the
Phoenician cities, older or newer, has left any traces of exploration
among Atlantic islands other than these or added any mythical
islands to maps or legends, unless through successors translating
into another language. The crowning achievement of the Phoeni-
cians, so far as we know, was the circumnavigation of Africa by
mariners in the service of Pharaoh Necho some 700 years before
Christ. This would naturally have brought them en route into
contact with the Canary and Cape Verde Islands, and they would
be likely to pass on to the Egyptians and Greeks a report of the
attributes of those islands partly embodied in names that might
adhere.

[Autolocus]

THE GREEKS AND ROMANS

We know that the Greeks of Pythias' time coasted as far
north as Britain and probably Scandinavia and had most likely
made the acquaintance still earlier of the Fortunate Islands
(two or more of the Canary group), similarly following downward
the African shore. Long afterward the Roman Pliny knew Ma-
deira and her consorts as the Purple Islands; Sertorius contem-
plated a possible refuge in them or other Atlantic island neigh-
bors; and Plutarch wrote confidently of an island far west of

Historical Library, Vol. i, Bk. 5, Ch. 2, p. 309.



THE NORSEMEN 5

Britain and a great continent beyond the sea where Saturn slept.
Other almost prophetic utterances of the kind have been culled
from classical authors, but they have mostly the air of specula-
tion. It cannot be said that the Greeks or Romans devoted
much energy to the remoter reaches of the ocean.

[Autolocus]

IRISH SEA-ROVING

Ireland was never subjectjto J^me^tjiQUgh. influenced by
Roman tracfeTand culture. From prehistoric times the Irish had
done some sea rovmg71isl:heir Imrama, or sea sagas, attest; and
this roving was greatly stimulated in the first few centuries of
conversion to Christianity by an abounding access of religious
zeal. Irish jnonjcs_seem^toj^y^settledm the end

(rf the^elghth century_and even to have sailed well beyoncflE
"There are good reasons~fo!T5eTteving that they had visited most
of the islands of the eastern Atlantic archipelagoes. We cannot
suppose that this rather reckless persistency ended there in such
a period of expansion. It is quite possible that we owe to this
trait the Island of Brazil, in the latitude of southern Ireland,
as an American souvenir on so many medieval maps (Ch. IV).
It is certain that the "Navigatio" of St. Brendan scattered St.
Brandianjslands, real or fanciful, over t!ie~bcean wastes of a cred-
ulous cartography (Ch. III).

[Autolocus]

THE NORSEMEN

A little later Scandinavians followed along the northern route,
finding convenient stopping points in the Faroes and Iceland,
discovered Greenland, and planted two settlements on its south-
western shore in the last quarter of the tenth century (Ch. VII).
Some of their ruins, a less number of inscriptions, and many frag-
mentary relics and residua are found, so that we can form a good
idea of their manner of life. Such as it was, it endured more than
four hundred years. To contemporary and slightly later geog-
raphy Greenland appeared most often as a far-flung promontory
of Europe, jutting down on the western side of the great water;

[Autolocus]

6 INTRODUCTION

but sometimes it was thought of as an oceanic island, with greater
or less shifting of location, and seems to be responsible for divers
mythical Green Islands of various maps and languages.

Less than a quarter of a century after their first landing the
Norse Greenlanders became aware of a more temperate coast line
to the southwest, the better part of which they called Vinland, or
Wineland, but all of which we now name America. Perhaps
Leif Ericsson brought the first report of it as the result of an
accidental landfall close to the year 1000 A. D. Not long after-
ward, Thorfinn Karlsefni with three ships and 160 people at-
tempted to colonize a part of the region. The venture failed, ow-
ing chiefly to the hostility of the Indians at the most favorable
point. The visitors, however, made the acquaintance of the
typical American Atlantic shore line of beach and sand dune
which stretches from Cape Cod to the tip of Florida with one or
two slight interruptions and one or two fragmentary minor
northward extensions. The Norsemen or some predecessor had
observed and named the three great zones of territory which
must always have existed. Among investigators there has been
general concurrence as to their discovery of Labrador and New-
foundland, to which most would add Cape Breton Island and
more or less of the coast beyond. It has appeared to me that they
made their chief abode in the New World on the shore of Passa-
maquoddy Bay behind Grand Manan Island and Grand Manan
Channel, with the racing ocean streams of the mouth of the Bay
of Fundy; and that they found this site inclement in winter and
tried to remove to a land-locked bay of southern New England
but were baffled and withdrew. My reasons have been pretty
fully set forth in "Early Norse Visits to North America." 7 For the
present it is enough to say that the discovered regions seem some-
times to have been thought of as a continuous coast line, some-
times as separate islands more or less at sea. But they did not
get upon the maps in any shape until several centuries later.

''Smithsonian Miscellaneous Collections. Vol. 59, No. 19, Washington, D. C.,
1913. See also: Recent History' and Present Status of the Vinland Problem, Ceogr.
Rev., Vol. ii, 1921, pp. 265-282.

[Autolocus]

MOORISH VOYAGES 7

MOORISH VOYAGES

The Moors who conquered Spain took up the task of Atlantic
exploration from that coast after a time. Its islands appear in
divers of the Arabic maps. In particular we know through
Edrisi, 8 the most celebrated name of Arabic geography, of the
extraordinary voyage of the Moorish Magrurin of Lisbon, who
set out at some undefined time before the middle of the twelfth
century to cross the Sea of Darkness and Mystery. They touched
upon the Isle of Sheep and other islands which were or were to
become notable in sea mythology. Perhaps these islands were
real, but they are not capable of certain identification now.
These Moorish adventurers seem to have reached the Sargasso
Sea and to have changed their course in order to avoid its im-
pediments, attaining finally what may have been one of the
Canary Islands, where they suffered a short imprisonment and
whence, after release, they followed the coast of Africa home-
ward. Edrisi about 1 154 wrought a world map in silver (long lost)
for King Robert of Sicily and also wrote a famous geography illus-
trated by a world map and separate sectional or climatic maps.
He devotes some space to Atlantic islands and their legends,
shows a few of them, and believes in twenty-seven thousand;
but the very few copies of his work which remain were made at
different periods and in different nations, and their maps dis-
agree surprisingly; so that it is not practicable to restore with
certainty what he originally depicted. He seems to have had at
least some acquaintance with the authentic island groups from
the Cape Verde Islands to the Azores and Britain. The fantastic
legends he appends to some of them do not seem to have greatly
affected the prevailing European lore of that kind.

[Autolocus]

8 Edrisi's "Geography," in two versions, the first based on two, the second on
four manuscripts, viz. : (i) P. A. Jaubert (translator) : Geographic d'Edrisi, traduite
de 1'Arabe en Francais, 2 vols. (Recueil de Voyages et de Memoires public par la
Socieie de Geographic, Vols. 5 and 6), Paris, 1836 and 1840; reference in Vol. 2,
p. 27; (2) R. Dozy and M. J. De Goeje (translators): Description de 1'Afrique et
de 1'Espagne par Edrisi: Texte arabe public pour la premiere fois d'apres les man.
de Paris et d'Oxford, Leiden, 1866.



8 INTRODUCTION

ITALIAN EXPLORATION

The Italians of the thirteenth century undertook similar ex-
plorations and temporarily occupied at least one of the Canary
Islands, Lanzarote, which still bears, corrupted, the name of its
Genoese invader, Lancelota Maloessel, of about 1470. On early
fourteenth-century maps and some later ones the cross of Genoa
is conspicuously marked on this island in commemoration of the
exploit. It was probably at this period that Italian names were
applied to most of the Azores and to other islands of the eastern
groups. A few of these names still persist, for example, Porto
Santo and Corvo; but others, after the rediscovery, gave way to
Portuguese equivalents or substitutes. Thus Legname was
translated into Madeira, and Li Conigi (Rabbit Island) became
more prettily Flores (Island of Flowers). About 1285 the Geno-
ese also sent out an expedition 9 "to seek the east by way of the
west" under the brothers Vivaldi, who promptly vanished with
all their men. Long afterward another expedition picked up on
the African coast one who claimed to be a survivor; and it is
probable that the Genoese expedition attempted to sail around
Africa but came upon disaster before it was far on its way. The
thirteenth- and fourteenth-century Italians undoubtedly added
many islands to the maps or secured their places there; but we
have no evidence that they passed westward beyond the middle
of the Atlantic.