Beside those legendary Atlantic islands that may cast some light on visits of white men to America before Columbus or have been at some time linked therewith by speculation or tradition—notably Antillia and its consorts, Brazil, Man or Mayda, Green Island, Estotiland and Drogio, the Island or Islands of St. Brendan, and the Island of the Seven Cities—there are numerous others, quite a swarm indeed, excusing Ptolemy’s and Edrisi’s extravagant estimate of 27,000. Sometimes, but not always, they are of more recent origin and are explainable in various ways.
Several are linked to the idea of volcanic destruction or seismic engulfment. Of course the colossal and classical instance of Atlantis comes first into mind, it being the earliest as well as in every way the most imposing. Most likely the well-known story, repeated, if not originated, by Plato, developed naturally, as we have seen, from the insistent need to account for the obstructive weedy wastes of the Sargasso Sea beyond the Azores and recurrent facts of minor cataclysms among them.
The next oldest instance, perhaps, is supplied by Ruysch’s map of 1508,289 an inscription on which avers that an island in the sea about midway between Iceland and Greenland had been totally destroyed by combustion in the year 1456. We do not know his authority for this startling announcement. The spot is where one would naturally look for Gunnbjörn’s skerries of the older Icelandic writings; and no one can find them now, unless they were, after all, but projecting points of the eastern Greenland coast. Also Iceland is at times tremendously eruptive; and this islet, or these islets, would not be far away. The assertion is not in itself incredible, but there seems no corroboration.
The “Sunken Island of Buss” presents a suggestion of engulfment on a more extensive scale. The whole episode is of rather recent date, Buss being the latest born of mythical or illusory islands, unless we except Negra’s Rock and other alleged and unproven apparitions of land on a very small scale, which may not have wholly ceased even yet. Buss is, at any rate, the one moderately large phantom map island the time and occasion of whose origin are securely recorded. For, as narrated by Best and published in Hakluyt’s compilation, on Frobisher’s third voyage (1578), one of his vessels, a buss, or small strong fishing craft, of Bridgewater, named Emmanuel, made the discovery. In his words:
The Buss of Bridgewater, as she came homeward, to the southeastward of Frisland, discovered a great island in the latitude of 57 degrees and a half, which was never yet found before, and sailed three days along the coast, the land seeming to be fruitful, full of woods, and a champaign country.290
Best must have had his information at second or third hand, with liberal play of fancy in the final touches on the part of his informant or himself. His was the first account published, but not long afterward appeared that of an eyewitness, “Thomas Wiars, a passenger in the Emmanuel, otherwise called the Busse of Bridgewater,” repeated in Miller Christy’s admirable little treatise on the subject.291 Wiars says they fell with Frisland (probably a part of Greenland) on September 8 and on September 12 reached this new island, coasted it for parts of two days, and considered it 25 leagues long. There was much ice near it. He gives no suggestion of fertility, woods, or fields.
The only other witnesses to the visual existence of the island, so far as recorded, were James Hall (probably by honest mistake) in 1606 and Thomas Shepherd (gravely distrusted) in 1671.292 Nevertheless an impressive insular figure grew up in the maps, bearing the name “Buss” to commemorate the vessel that first found it. In some instances it was made a very large island indeed. Shepherd’s map, reproduced herewith (Fig. 24), was accompanied by a brief descriptive narrative which may be attributed to a fancy for yarning, with no strong curb of conscience on the fancy. Buss remained an accepted figure of geography for considerably more than a century.
Quite naturally, however, the efforts of reliable searchers failed to find this island again, for it was not really there. A theory of cataclysm seemed more acceptable than to discard outright what so many maps, books, and traditions had attested. Van Keulen’s chart of 1745293 led the way with the inscription “The submerged land of Buss is nowadays nothing but surf a quarter of a mile long with rough sea. Most likely it was originally the great island of Frisland.” So the name “Sunken Land of Buss” passed into general use with geographic sanction. After much disturbance of mariners’ and cartographers’ minds not only the phantom island but its legacy, the supposed line of breakers and dangers, vanished altogether from the records. There is no “Buss” to be found on maps after about the middle of the nineteenth century, though the preceding hundred years had been prolific in them. Probably we must suppose a later date for the cessation of current mention of the sunken land of that name, in recognition of what, according to belief, once had been but existed (above water) no longer.
Indeed, even after the opening of this twentieth century the same hypothesis has revived,294 with scientific support of a submarine range in 53° N. and 35° W., really ocean-bottom mountains 8,000 feet high between Ireland and Newfoundland, reported upon in 1903 by Captain de Carteret of the cable ship Minia. They are not on the same spot and would still require a great lift to reach the surface. Of course their past sinking is not impossible, but there is no need to explain Buss by cataclysm any more than Mayda or Brazil Island, Drogio or Icaria.
Somewhat allied by nature to these reported isles of destruction and disappearance are the islands of imported diabolism, appearing on maps now and then through the centuries. Bianco’s “The Hand of Satan” (1436295; Fig. 25), if correctly translated (see Ch. X, p. 156), is probably the first to present this quality. He locates the sinister island well to the southward; but the most pictorial appearance is Gastaldi’s (for Ramusio) “Island of Demons,”296 with its eager and capering imps at the bleak and savage northern end of Newfoundland. The preferred site, however, would seem to be yet a little farther north. Ruysch, in the map referred to above, which announces the burning up of Gunnbjörn’s skerries, exhibits two Insulae Demonium near the middle of the dreaded Ginnungagap passage between Labrador and Greenland. There is no suggestion of volcanic action in their case, and it does not appear that any real islands occupied the spot. The reason for the delineation and the name is still to seek.
The map of 1544, attributed to Sebastian Cabot,297 makes a single island of them, “marked Y. de Demones”, and brings it nearer the eastern front of Labrador below Hamilton Inlet. Agnese298 in the same century enlarges it greatly but still keeps it just off the Labrador coast. The Ortelius map of 1570299 (Fig. 10) shows the insular haunt of devils, plural again in form and name, but retains approximately the site chosen by Cabot. Mercator’s world map of 1569300 keeps the islands plural beside the upper tip of Newfoundland, approximating Gastaldi’s position. There seems to have been a pronounced and general concurrence of belief in diabolical evil in the northeastern coast of America, perhaps because it is there that the Arctic current brings down its tremendous freight, and tempests are at their wildest, and all barrenness and bleakness at their worst.
Much farther south, on the lines followed by Columbus and his Latin successors and in the tracks of vessels plying between the eastern Atlantic archipelagoes and the West Indies, what may be considered as a contrary impulse—that of exultant religious enthusiasm—came into play in island naming. The Island of the Seven Cities (Ch. V) will be recalled but needs no further consideration here. St. Anne, La Catholique, St. X, and Incorporado (in the sense of Christ’s Incarnation) are among the more conspicuous instances. The second-named was always in low latitudes. It occurs in the latitude of the tip of Florida, in mid-Atlantic in the Desceliers map of 1546301 (Fig. 9); also as “La Catolico” on Portuguese maps, with similar situation. Desceliers shows Encorporade (Incorporado) about east of Cape Hatteras and south of western Newfoundland; but he also has Encorporada Adonda not far from Nova Scotia. Thomas Hood (1592)302 makes a wild and unenlightened transformation of Incorporado to “Emperadada” and puts it about opposite the site of Savannah, but not so far east as the considerable outjutting of the coast which must be meant for Cape Hatteras and its neighborhood. However, this location is not very different from that usually given it. Desceliers has two islands marked St. X, one being in the longitude of St. Michaels and latitude of Bermuda; the other in the longitude of eastern Newfoundland and latitude of the Hudson. In about the same latitude as the latter, and more than half way between it and the Azores, an island called St. Anne is shown. There seems nothing real to prompt the derivation of these religiously named islands. Perhaps they are merely the offspring of optical delusion, fancy, and fervor.
On the other side of the Atlantic the much earlier map island Daculi must be reckoned as of kin to them, since its map legends deal with beneficent wonder working or magical medical aid, and its name may be identical with or have originated the saintly one which still denotes an outlying Hebridean island. Though less renowned than the island of Brazil and less significant, Daculi shares with it the record for first appearance of mythical islands on portolan maps.
Dalorto’s map of 1325303 (Fig. 4) already indicated as the earliest one of much interest in this special regard, presents many islands of familiar or unfamiliar names near Ireland and Scotland. Nobody can mistake the rightly located Man, Bofim, and Brascher (the Blaskets). Insula Sau must be Skye, though with the outline of the Kintyre peninsula. Sialand seems to be Shetland. Tille may be Orkney displaced. Galuaga or Saluaga probably stands for the main body of the Long Island (Harris, Lewis, etc.) of the outer Hebrides. Bra is no doubt Barra and has generally been thus accepted, though out of line with Galuaga and too far eastward. Brazil, as already reported, is naturally farther at sea opposite Brascher. Finally our subject for present consideration, Daculi, lies off the northwestern corner of Ireland, north of Brazil Island and west of Bra, with which last it has in later maps a curious legendary association. With Insula de Montonis, as Brazil is also called on Dalorto’s map, it may be linked in another way by their Italian names, for Daculi seems capable of that derivation, “culla” being “cradle” in that language, plural “culli,” easily modified to “culi” by careless speech or writing. The introductory preposition “da” in one use has an especial relation to nativity; thus Zuan da Napoli means John born at Naples, that is John of Naples in this sense. The blending of preposition and noun in one word, “Daculi,” is no more than sometimes happened on the maps to the article and noun “Li Conigi,” the Rabbit Island, making it “Liconigi,” now long known as Flores. This explanation would interpret Daculi as the “Island of the Cradles,” or “Cradle Island.” Some other derivation may indeed possibly be as defensible; but it should be borne in mind that Italian traders ranged very early up and down the Irish coast, and that name would curiously coincide with the tradition at least afterward current concerning the island.
To review a few later but still very early maps:—Dulcert, 1339,304 shows some irrelevant changes farther north and east; but his Hebridean islands repeat very nearly the form given them by Dalorto (believed by many to be the same man), and there is no significant change in Bra or Daculi, though the first syllable of the latter becomes Di.
The Atlante Mediceo, of 1351,305 makes more changes than Dulcert among these islands and leaves unnamed the one which by position seems meant for Bra, or Barra. Daculi is largely expanded and named Insul Dach indistinctly.
The Pizigani map of 1367306 (Fig. 2) modifies many names. Daculi becomes Insuldacr in one word; but its place remains nearly as in Dalorto’s map, though most of the other islands are drawn closer to Ireland, so that Bra is nearly stranded thereon. A line of inscription seems to relate to Bra—“Ich sont ysula qu—[possibly pronominal abbreviation] abitabi honõ quõ morit may.” Perhaps some of these words should be read differently, and “abitabi” needs some recasting. I will not attempt to interpret but should infer that Bra had its troubles. They do not seem to have extended to Daculi.
Pareto’s fine map of 1455307 (Fig. 21) applies the following more extended and significant legend to Daculi: “Item est altera insulla nomine Bra in qua femine que in insulla ipsa habitant non pariuntur sed quando est eorum tempus pariendi feruntur foras insulla et ibi pariuntur secundum tempus.” From this we may gather that the outer island Daculi was believed to afford especial aid in childbearing to women carried thither after being baffled on the inner island Bra, and we see readily the appositeness of the name “cradle” applied to the former. Beccario’s map of 1435308 (Fig. 20), though without the legend, had already adopted in “Insulla da Culli” almost exactly the form of the name which we have divined, with apparently that meaning.
St. Kilda seems to me the most plausible original for Daculi that has been suggested. It is true that Barra is actually south of the parallel of latitude of that most lonely western sentinel of the Hebrides, and there is no obvious link of relation between them. Also the rock islet of North Barra is about as far above it, equally unconnected and not likely ever to have maintained much population. But so simple a misunderstanding on the part of the old cartographers would be no more than what happened to them all the time, and exact identity of latitude is unimportant. There is, in fact, no land on the site given Daculi in any of these old maps; and Bra, as noted, is absurdly out of place for Barra. How the tradition grew up we do not know. Perhaps it was some tale picked up by coasting Italian traders, partly misunderstood and passed on by them to the map-makers at home. St. Kilda, lost in the mists and mystery of the Atlantic, of holy name and miracle-working associations, and out of touch with most tests of reality, seems a likely place to be linked to some less abnormal island by a fanciful contribution of saintly white magic, a rumor originating nobody knows how.
On the western side of the Atlantic there are divers instances of island names given of old—sometimes with considerable changes of location, area, or outline, or of all three—to regions which we know quite otherwise. Some of these have been dealt with extensively already. Greenland has a lesser neighbor, Grocland, on its western side in divers sixteenth-century maps; which I take to be a magnified presentation of Disko or possibly a reflection of Baffin Land brought near. It appears conspicuously in Mercator’s map of the Polar basin (1569),309 the Hakluyt map of 1587 illustrating Peter Martyr,310 and the map of Mathias Quadus (1608).311
This is not the place to enlarge on the Helluland, Markland, and Vinland of the Norsemen beginning with the eleventh century, as this theme has been dealt with elsewhere.312 But they were often thought of as islands, as shown by the notice of Adam of Bremen. Perhaps there was never any great clearness of conception as to extent or form. But in a general way they may be identified respectively with northern Labrador, Newfoundland, and the warmer parts of the Atlantic coast. Great Iceland, or White Men’s Land, seems also to have been understood as what we should now call America. Eugène Beauvois located it conjecturally about the mouth of the St. Lawrence River.313 Dr. Gustav Storm, on the other hand, thought it was merely Iceland misunderstood.314
Perhaps the latter explanation is the best yet given of the mysterious island Scorafixa, or Stokafixa, in Andrea Bianco’s map of 1436.315 It has sometimes been understood as Newfoundland, which bore long afterward the name Bacalaos, the equivalent in a different tongue of the northern “stockfish,” our codfish. But it would naturally be freely applied to any island in rather high latitudes which was conspicuous for that fishery, and Stokafixa seems near of kin to Fixlanda, which figures on divers maps as a combined suggestion of Iceland and the imaginary Frisland but with geographical features mainly borrowed from the former. The first-named identification may be tempting as establishing another pre-Columbian discovery of America, but it quite lacks corroboration; and Iceland was a great center of codfishery, distributing its name and attributes rather liberally in legend and on the maps. Humboldt incidentally mentions “l’île des Morues (île de Stockfisch, Stokafixa)” on the seventh map of the atlas of Bianco, 1436. I do not clearly make out the name on T. Fischer’s facsimile reproduction;316 but from position and appearance the island seems meant for Iceland.
The Grand Banks and other banks of Newfoundland, with the Virgin Rocks and perhaps other piles or pinnacles rising from that bed nearly to the surface so as to be uncovered in some tides; Sable Island, a rather long way offshore; Cape Breton Island and fragments of the main shore—may be held responsible for some map islands such as Arredonda and Dobreton, Jacquet I., Monte Christo, I. de Juan, and Juan de Sampo.
There are still other islands mostly north of the latitude of Bermuda and between it and the Azores or northeastern America, but far at sea, of which one can make little, except as probably complimenting some pilot, skipper, or other individual, or commemorating some incident which has nevertheless been generally forgotten. Thus Negra’s Rock, which has hardly ceased to appear on the maps, does not really exist but may keep us in mind, by its rather sinister and mythical sound, that a certain Captain Negra once thought he saw something solid in the great liquid and reported accordingly. Of such origin, perhaps, are I. de Garcia, Y Neufre, Y d’Hyanestienne, Lasciennes, and divers others scattered over various maps and offering no promise of reward for hunting down their pedigrees or history. All these distinctly post-Columbian islands are quite too recent and casual to throw any light on the earlier historically and geographically significant “mythical islands” or on what these reveal.