NO. IQ NORSE VISITS TO NORTH AMERICA BABCOCK 17 



panions, were generally accepted by the conjecture or tradition of the 

 fourteenth century as The Fortunate Islands and especially The 

 Fortunate Isles of St. Brandan. This identification was afterwards 

 forgotten, but the memory lingered that at least one island had borne 

 his name, and we find it reappearing, in random fashion, here and 

 there about the ocean at points where no island should be a very 

 elusive " He fantastique," though not the only one nor the most sig- 

 nificant ; for " the mythical islands " are sown most liberally over the 

 maps of five centuries. Thus Brazil, on a French map of 1754 holds 

 nearly the same direction from Limerick as Dalorto gave it about 430 

 years before. Mayda (Asmayda) is even more persistent, for I find 

 it in the old and proper latitude, opposite northern France, on a relief 

 map, copyrighted in the United States in 1906. 



As map-makers have generally followed explorers, with only a 

 little toning down and conjectural improvement, we may safely 

 take every additional island of the map as representing at least 

 one voyage or the report of one. We know how the very dubious 

 disclosures of the Zeni and the indubitable discoveries of the fifteenth 

 century got into geography, though the former have since melted 

 away. Also we can see how the medieval cartographers built up, 

 item by item, a true island-showing for the eastern side of the Atlantic, 

 so that even the 1351 map already cited, 1 has not only all the Canaries, 

 but all their names, as now in use, with the single exception of 

 Tenerifife. The islands which have not held their place in maps of the 

 best authority are almost all islands out of place and duplicated, 

 like the Island of St. Brandan, or bits of some more extended and 

 more distant coast line similarly misunderstood. Thus the Sunken 

 Land of Bus, named after one of Frobishers ships ' and long a dis- 

 quiet to the mariner, since it could never be found again, is now 

 generally recognized as a part of Greenland, which appeared un- 

 expectedly before him when he was somehow off his reckoning. 

 Several other and better known " mythical islands " are inadequately 

 accounted for by any theory which does not cross the Atlantic. 



In form and direction Antillia and Brazil are quite as constant as 

 the Canaries, and more so than the Azores, of the early maps ; which 

 may show conviction arising from some previous precise narrative. 

 Antillia, at its first appearance, is a large, elongated, rectangular, 

 quadrilateral island with four indentations in its eastern side, three in 

 its western side, each in two or three lobes, also a greater one at its 

 southern end, all carefully delineated as if by survey; and it so re- 

 mains, on nearly all the pre-Columbian maps. Sometimes this form 



! M. D'Avezac: Discoveries of the Middle Ages, p. 42. 

 -Or possibly after one of his officers. 



