The mythical islands of the Atlantic (les îles fantastiques) on the old maps have had divers origins, instructive to study. Perhaps only one of them derives its name and being directly from a real human episode of a twilight period in history.
When the Moors descended on Spain in 711, routed King Roderick’s army beside the Guadalete, and rapidly overran the Iberian Peninsula, it was most natural, indeed nearly inevitable, that some Christian fugitives should continue their flight from the seaboard to accessible islands already known or rumored, or even desperately commit themselves in blindness to the remoter mysteries of the ocean. Such an event would afford a fabric for the embroidery of later fancy. A part of this has been preserved by record; and it is curious to watch the development of the story, which takes several forms, not differing widely, however, one from another.
When Pedro de Ayala, Spanish Ambassador to Great Britain, found occasion in 1498 to report English exploring activities to Ferdinand and Isabella, he wrote:
The people of Bristol have, for the last seven years, sent out every year two, three, or four light ships (caravels) in search of the island of Brasil and the seven cities.115
There is indeed one well-attested voyage of 1480 conducted by well-known navigators, seeking this insular Brazil, and it was not the earliest.
The first appearance of that island thus far reported, as we have seen in the preceding chapter, is on the map of Dalorto116 (dated 1325; Fig. 4) as a disc of land well at sea, westward from Hibernian Munster; but the Catalan map of 1375117 (Fig. 5) and at least one other118 turn the disc into a ring surrounding a body of water which is studded with small islands—apparently nine in the Catalan map photographically reproduced by Nordenskiöld, though Dr. Kretschmer draws seven on the other. These miniature islands have sometimes been thought119 to represent the seven cities of the old legend; but islets are not cities, and there seems no reason why each city should require an islet. However, the coincidence of number, exact or approximate, is suggestive.
Antillia (variously spelled) was a home for the elusive cities more favored than Brazil by cartography and tradition. In 1474 Toscanelli, a cosmographer of Florence, being consulted by Christopher Columbus as to the prospects of a westward voyage, sent him a copy of a letter which he had written to a friend in the service of the King of Portugal. Its authenticity has been questioned, but it is still believed in by the majority of inquirers and may be accepted provisionally. In it occurs this passage:
From the island Antilia, which you call the seven cities, and whereof you have some knowledge, to the most noble island of Cipango [Japan], are ten spaces, which make 2,500 miles.120
The name Antillia had appeared on the maps much earlier. As Atilae, or Atulae, it is doubtfully found in an inscription on that of the Pizigani (1367;121 Fig. 2), identifying a “shore,” not drawn, on which a colossal statue of warning had been erected. The location seems to be somewhere in the region where Corvo of the Azores should appear.
We meet the island name, for the first time unmistakably, on the map of Beccario (Becharius) of 1435122 (Fig. 20). It is applied to the chief of a group of four large islands, comparable to nothing actually in the western Atlantic except the Greater Antilles, or three of them with Florida (Bimini). They are collectively designated “Insulle a Novo Repte”—the “Newly Reported Islands.” Antillia itself is shown as an elongated quadrilateral having its sides indented by seven two-lobed bays of identical form, beside another and larger bay in the southern end. Several subsequent maps repeat the delineation with little change, and the map of Benincasa (1482;123 Fig. 22) supplies local names for the bays or the regions adjoining excepting only the lowest but one on the eastern side, which bay is opposite the middle of the island name Antillia. The other names as read by Dr. Kretschmer are Aira, Ansalli, Ansodi, Con, Anhuib, Ansesseli, and Ansolli. It will be observed that five of them borrow the first syllable of Antillia. Nobody has explained these names, and they seem mere products of linguistic fancy. But again the coincidence in number is impressive, although somewhat offset by the fact that the next largest island in the group, Saluaga, has a similar arrangement of five bays of like form and carries the names, similarly applied, of Arahas, Duchal, Imada, Nom, and Consilla. They can hardly be extra bishops’ towns. At least we are in the dark about them. The anonymous map sometimes attributed to 1424 and preserved at Weimar124 shows in photographic copy traces of names, or at least letters, on the part of Antillia which it represents. Its true date is believed to be about that of Benincasa’s map above cited. But the markings do not seem to be identical and are very meager.
However, there can be no doubt of Toscanelli’s meaning at an earlier date in the passage quoted. The same is true of Behaim’s globe (1492), though he discards the accepted form of Antillia. He appends a long inscription, translated by Ravenstein as follows:
In the year 734 of Christ, when the whole of Spain had been won by the heathen (Moors) of Africa, the above island Antilia, called Septe citade (Seven cities), was inhabited by an archbishop from the Porto in Portugal, with six other bishops, and other Christians, men and women, who had fled thither from Spain, by ship, together with their cattle, belongings, and goods. 1414 a ship from Spain got nighest it without being endangered.125
Again, in Ruysch’s map of 1508 there is “a large island in the middle of the Atlantic Ocean between Lat. N. 37° and 40°. It is called Antilia Insula, and a long legend asserts that it had been discovered long ago by the Spaniards, whose last Gothic king, Roderik, had taken refuge there from the invasion of the Barbarians.”126
Ferdinand Columbus, living between 1488 and 1539, says that some Portuguese cartographers had located
Antilla ... not ... above 200 leagues due west from the Canaries and Azores, which they conclude to be certainly the island of the seven cities, peopled by the Portuguese at the time that Spain was conquered by the Moors in the year 714. At which time they say, seven bishops with their people embark’d and sailed to this island, where each of them built a city; and to the end none of their people might think of returning to Spain, they burnt the ships, tackle and all things necessary for sailing. Some Portuguese discoursing about this island, there were those that affirmed several Portuguese had gone to it, who could not find the way to it again.127
He relates particularly how “in the time of Henry infant of Portugal [perhaps about 1430], a Portuguese ship was drove by stress of weather to this island Antilla.” The crew went to church with the islanders but were afraid of being detained and hurried back to Portugal. The Prince heard their story and ordered them to return to the island, but they escaped from him and were not found again. It is said that of the sand gathered on Antillia for the cook room a third part was pure gold.
Galvano tells of a still later visit; or possibly it is only another version of the same:
In this yeere also, 1447, it happened that there came a Portugall ship through the streight of Gibraltar; and being taken with a great tempest, was forced to runne westwards more then willingly the men would, and at last they fell upon an Island which had seven cities, and the people spake the Portugall toong, and they demanded if the Moors did yet trouble Spaine, whence they had fled for the losse which they received by the death of the king of Spaine, Don Roderigo.
The boateswaine of the ship brought home a little of the sand, and sold it unto a goldsmith of Lisbon, out of the which he had a good quantitie of gold.
Don Pedro understanding this, being then governour of the realme, caused all the things thus brought home, and made knowne, to be recorded in the house of justice.
There be some that thinke, that those Islands whereunto the Portugals were thus driven, were the Antiles, or Newe Spaine.128
The Portuguese historian Faria y Sousa has yet another version. According to Stevens’ translation:
After Roderick’s defeat the Moors spread themselves over all the province, committing inhuman barbarities. * * * The chief resistance was at Merida. The defendants, many of whom were Portuguese, that being the Supreme Tribunal of Lusitania, were commanded by Sacaru, a noble Goth. Many brave actions passed at the siege, but at length there being no hopes of relief and provisions failing, the town was surrendered upon articles. The commander of the Lusitanians, traversing Portugal, came to a seaport town, where, collecting a good number of ships, he put to sea, but to which part of the world they were carried does not appear. There is an ancient fable of an island called Antilla in the western ocean, inhabited by Portuguese, but it could never yet be found, and therefore we will leave it until such time as it is discovered, but to this place our author supposes these Portugals to have been driven.129
It is plain that Captain Stevens paraphrases with comments rather than translates. The original130 avers that the fugitives made sail for the Fortunate Islands (the Canaries), in order that they might preserve some remnants of the Spanish race, but were carried elsewhere. It also specifies that the legendary island which they are supposed to have reached is inhabited by Portuguese and contains seven cities—tiene siete cividades.
This last account lacks positive mention of the emigrating bishops and for the first time names a definite though rather remote goal as aimed at by their effort. But the movement from Merida is well accounted for, and a trusted military commander would seem a natural leader for such an enterprise of wholesale escape. The bishops, implied by the seven cities, might well gather to him at Oporto or be picked up on the way. On the whole it seems the most easily believable version of the story; though of course it does not necessarily follow that they really chose any land so remote as Teneriffe and its neighbors—if they knew of them—for a new abiding place. Of course the continuance of Portuguese language and civilization and the persistence of seven isolated towns through so many centuries must be ranked with the auriferous sands of Antillia as late products of the dreaming Iberian brain.
The citations thus far given identify the Island of the Seven Cities with some legendary, but generally believed-in patch of land afar out in the ocean—sometimes with the Island of Brazil, more often with Antillia. But the earliest of them dates six or seven centuries after the supposed fact, and it may well be that a distinction was made at first, which became lost afterward by blending. In a still later stage of development the name of the Seven Cities becomes separate and strangely migratory, not avoiding even the mainland. We know, for instance, what power the Seven Cities of Cibola had to draw Coronado and his followers northward through the mountains and deserts of our still arid Southwest until all that was real of them stood revealed as the even then antiquated and rather uncleanly terraced villages of sun-dried brick which are picturesquely familiar on railway folders and in the pages of illustrated magazines.
But this was not the only part of North America on which the romantic myth alighted. The British Museum contains in MS. 2803 of the Egerton collection an anonymous world map,131 (Fig. 8), forming part of a portolan atlas attributed by conjecture to 1508, which shows, somewhat as in La Cosa’s map of 1500, the Atlantic coast distorted to a nearly westward trend, with the Seven Cities (Septem Civitates), represented by conventional indications of miters, scattered along a seaboard tract from a point considerably west of “terra de los bacalos” and the Bay of Fundy to a point nearly opposite the western end of Cuba. The cartographer’s ideas of geography were exceedingly vague, but apparently he conceived of Portuguese episcopal domination for the coastal country between lower New England and Florida as we know them now. Perhaps, however, he merely meant to set down his cities somewhere on the eastern shore of temperate North America and has strewn them along at convenience.
Incidentally, this map is also interesting as one of a few which inscribe Antillia, with slight changes of orthography, on some part of the mainland of South America. In this instance “Antiglia” occupies a tract of the northwestern coastal country apparently corresponding to contiguous portions of Colombia, Ecuador, and Peru.
But the Island of the Seven Cities appeared as such on other maps and by this name only. Perhaps its most salient showing is on Desceliers’ fine map of 1546132 (Fig. 9), that entertaining repository of isles which are more than dubious and names which are fantastic. He presents it off the American coast about a third as far as the Bermudas and midway from Cape Breton to the Bay of Fundy. The size is considerable, the outline being deeply embayed on several sides and hence very irregular, almost as much so as Celebes. Two islets lie near two of its projecting peninsulas. It bears a brief inscription giving the name Sete Cidades and indicating that it belongs to Portugal.
This choice of location would have been more venturesome a century later. In 1546 there had been some exploring and much fishing in these waters but no determined settlement near them, and they were hardly yet familiar. However, the Ortelius map of 1570133 (Fig. 10), and the Mercator map of 1587134 find it more prudent to move this island farther south and farther out to sea, reducing its area, but retaining its traditional name. Not long after this, except for a local name on St. Michaels of the Azores, the Seven Cities disappear from geography.
The exception noted is well worth considering. Just as Terceira retains her medieval name of Brazil to designate one headland, St. Michaels has still its valley of the Seven Cities. Brown’s guidebook presents the fact very casually: “St. Michaels. Ponta Delgada. Brown’s Hotel. About ten people. Among the chief sights are the lava beds coming from Sete Cidades.... At Sete Cidades, which is worth a visit, there is a great crater with two lakes at the bottom, one of which appears to be green, the other blue.”135
This naïve incuriousness in the presence of something so significant of course has not been shared by a different order of observers. Buache136 found here as he thought the genuine and only Seven Cities of the legend. Humboldt137 opposed this view with a reminder of the Seven Cities of Cibola. But it is fair to remember that New Mexico was quite impossible for the Portuguese of 711 or thereabout, whereas St. Michaels Island offered an accessible and tempting place of refuge. The name could not have been derived from settlement in the former; but it might really be derived from settlement in the latter. Granting that the fugitives might not be able to maintain themselves there in safety for many years after the Arabs had begun their tentative and always uneasy incursions into the western Sea of Darkness, it still may be that the town or towns of this hidden island valley might endure long enough and seem imposing enough and be visited often enough by Christians from the mainland to supply the nucleus of the most picturesque and adventurous of legends; and this tale might follow any later migration into the unknown, or survive and find new abiding places for the name and fancy long after the original colony—archbishop and bishops and congregations, military commanders, and mailed soldiery—had all been somehow destroyed or had melted apart and drifted away. All that remains certain is the continued presence of the name of the Seven Cities on that spot.
Some ruins are said to have marked it formerly, but very little is visible now, if we may trust the following description by an intelligent visitor in the middle of the last century:
Emerging from these sunken lanes, so peculiar to the island of St. Michael’s, we come to the green hills which border the village and the valley of the Seven Cities.... From these dull evergreen mountains, stretching before us without apparent end, we speedily had an unexpected change. Suddenly the mountain track up which we were climbing ended on the edge of a vast precipice, hitherto entirely concealed, and at a moment’s transition disclosed a wide and deeply sunk valley with a scattered village and a blue lake. The hills which hemmed them in were bold and precipitous, tent-shaped, rounded and serrated. Others swept in soft and gentle lines into a little plain where the small village was nestled by the water side. The lake was of the deepest blue and so calm that a sea bird skimming over its surface seemed two, so perfect was its image in the water. The clouds above were floating in this very deep lake, and the inverted tops of the hills on every side were perfectly reflected in its bosom. A few women on the shore seemed rooted there, so steady were their reflections in the water, and the cattle standing in the shallows stood like cattle in a picture.... The sides slope gradually from this part of the valley into the level ground where the village stands. It is a small collection of cottages, without a church or a wineshop or a store of any kind, and at the time I entered it was enveloped in clouds of wood smoke which rose from the fires used in the process of bleaching cloth. This and clothes washing are the chief occupations of the villagers....
A portion of the lake is separated from the larger one by a narrow causeway. It is singular to notice the difference made in the two pieces of water by this small embankment; for, while the large lake is clear and crystalline, this is thick, green, and muddy, and as gloomy as the Dead Sea, with no clouds or birds or bright sky reflected in it.138
Perhaps a little excavating archeology might not be amiss in the neighborhood of the causeway and the green dead lakelet. But at least it is satisfactory to have a good external account of the only site in the world, so far as I know, which still bears the legendary name. As elsewhere used, this name has certainly wandered widely and been affixed to many places. Whether any of these represent real refuges of the original emigrants or their descendants or others like them no one can quite certainly say; but there is no evidence for it, and the probabilities are against it. Certainly no Spanish nor Portuguese community, of Moorish or of any pre-Columbian times, established itself in western lands for any great period to make good the aspiration of the fugitives of Merida.